Credit Derivatives and the Sovereign Debt Restructuring Process

Pierre-Hugues Verdier

LL.M. Paper International Finance Seminar

Professors Hal S. Scott and Howell E. Jackson Harvard Law School

April 27, 2004

TABLE OF CONTENTS INTRODUCTION .................................................................................................................................................1 I. THE SOVEREIGN DEBT RESTRUCTURING DEBATE ..........................................................................4

A. GENERAL BACKGROUND: CRISES AND RESTRUCTURINGS ............................................................................4 1. The 1980s Sovereign Debt Crisis ..............................................................................................................4 2. The Secondary Market and the Brady Plan...............................................................................................6

B. THE HOLDOUT PROBLEM ...............................................................................................................................7 1. Origins of the Holdout Problem................................................................................................................7 2. Sovereign Debtors in U.S. Courts .............................................................................................................8 3. Towards the Nightmare Scenario?..........................................................................................................18 4. The 1990s Currency Crises and the Role of the IMF ..............................................................................22

C. REFORM INITIATIVES ...................................................................................................................................25 1. A Market-Based Solution: Collective Action Clauses ............................................................................25 2. An Institutional Solution: An International Bankruptcy Court? ............................................................29 3. Another View: The Case for Creditors’ Rights ......................................................................................31

II. CREDIT DERIVATIVES .............................................................................................................................34 A. CREDIT DEFAULT SWAPS.............................................................................................................................34 B. THE MARKET FOR CREDIT DERIVATIVES .....................................................................................................39

1. Usage of Credit Derivatives ....................................................................................................................39 2. Sovereign Credit Derivatives ..................................................................................................................42 3. Potential Market Failures .......................................................................................................................44

C. THE DOCUMENTATION OF CREDIT DERIVATIVES.........................................................................................47 1. ISDA Documentation...............................................................................................................................47 2. Credit Events and the Restructuring Debate...........................................................................................51

III. CREDIT DERIVATIVES AND THE RESTRUCTURING PROCESS .................................................58 A. POST-DEFAULT RESTRUCTURINGS...............................................................................................................58 B. PRE-DEFAULT RESTRUCTURINGS.................................................................................................................59

1. The Exchange Offer Framework .............................................................................................................59 2. Exchange Offers under the 1999 ISDA Definitions.................................................................................62 3. Effect on Incentives .................................................................................................................................64 4. Amendments to the ISDA Definitions ......................................................................................................70

C. COLLECTIVE ACTION CLAUSES....................................................................................................................72 1. Collective Action Clauses and the ISDA Definitions...............................................................................72 2. Do Collective Action Clauses Make the Obligations Contingent?..........................................................76

D. THE SDRM..................................................................................................................................................79 E. SOVEREIGN DEBT AND GLOBAL FINANCIAL STABILITY ...............................................................................81

1. Market Concentration and Systemic Risk................................................................................................81 2. Excessive Lending in Emerging Markets ................................................................................................83

CONCLUSION ....................................................................................................................................................84 APPENDIX: ISDA RESTRUCTURING DEFINITIONS...............................................................................88 BIBLIOGRAPHY................................................................................................................................................90

INTRODUCTION

Two of the most significant recent developments in the global financial system have

been the rise of a large secondary market for emerging market bonds, and the rapid expansion

of financial derivatives. The former took place in the wake of the 1980s Latin American debt

crisis and the long series of debt restructurings that ensued. Despite the continuing economic

troubles faced by these countries, much of their loan indebtedness to foreign banks was

eventually converted to widely-traded bonds under the Brady Plan. This conversion catalyzed

the development of a vast public market for the debt of ‘emerging markets,’ which now cover

much of the developing world.

This market, however, has been beset by serious difficulties. The recent defaults of

Russia and Argentina, which led to widespread economic misery and to serious social and

political unrest,1 highlighted the inadequacies of the multilateral bail-out approach developed

in the 1990s. In addition, the increasing dispersion of sovereign private debt further

complicates crisis management and debt restructuring efforts. These difficulties have been

compounded, in recent years, by a series of court rulings that eliminated the principal

defenses on which sovereign debtors had relied to protect their restructuring efforts against

legal action by creditors.

1 See, e.g., Edward Alden, Thomas Catan & Richard Lapper, Argentina’s President Quits; At Least 14 Killed in Riots as Country Falls into Chaos; Leadership Battle Breaks Out, FIN. TIMES, Dec. 21, 2001, at 1; Thomas Catan & Mark Mulligan, Argentina in Dollars 155bn Debt Default; Biggest Suspension Of Payments In History; ‘Third Currency’ To Be Created To Kick-Start Economy, FIN. TIMES (London ed.), Dec. 24, 2001, at 1; Thomas Catan, Hope – and Everything Else – Is Running Out in Buenos Aires: The Economic Crisis Is Swelling the Ranks of the Poor, FIN. TIMES (London ed.), Feb. 2, 2002, at 22; Betsy McKay, Moscow Journal: From Worry to Worse – Revisiting a Family in Russia 7 Years Later, WALL ST. J., Sept. 28, 1998, at A1.

- 1 -

The financial derivatives markets have also experienced stunning expansion. One of

the fastest-growing categories of instruments have been credit derivatives. 2 These

instruments, in their simplest form, are a promise from the seller to the buyer to make a

predetermined payment when an adverse credit event affects the entity to which the derivative

relates.3 Virtually inexistent as recently as ten years ago, this market has grown from an

aggregate outstanding notional amount of $893 billion at the end of 2000 to more than $1.9

trillion at the end of 2002. This trend is expected to persist, with some estimates projecting a

total of $4.8 trillion outstanding in 20044. Sovereign credit derivatives, traded over-the-

counter in New York and London, form a significant segment of this rapidly expanding

market5. Today, investors can buy credit protection on the external debt of Brazil, Mexico or

the Philippines, among others.

Despite the increasing importance of the sovereign credit derivatives market, and

some hints on the part of policy-makers that their use may further complicate restructuring

efforts,6 the possible effects of this development on global financial stability have not been

explored in detail. In this paper, I will review the existing private legal regime governing

credit derivatives and its interaction with the sovereign debt restructuring process. As will be

seen, the standard contract terms governing credit derivatives, if properly interpreted by

2 See Charles Batchelor, Credit Default Swaps Join Booming Derivatives Line-Up, FIN. TIMES, Feb. 11, 2004, at 26. 3 This entity is not normally a party to the transaction. 4 See BRITISH BANKERS’ ASSOCIATION (BBA), CREDIT DERIVATIVES REPORT 2001/02 (2002); see also INTERNATIONAL SWAPS AND DERIVATIVES ASSOCIATION (ISDA), MARKET SURVEY: HISTORICAL DATA (2003). 5 See COMMITTEE ON THE GLOBAL FINANCIAL SYSTEM (CGFS), CREDIT RISK TRANSFER (2003) [hereinafter CREDIT RISK TRANSFER]; Frank Packer & Chamaree Suthiphongchai, Sovereign Credit Default Swaps, BIS Q. REV., Dec. 2003, at 79. 6 See ANNE O. KRUEGER, A NEW APPROACH TO SOVEREIGN DEBT RESTRUCTURING 8 (2002) [hereinafter A NEW APPROACH] (“[Free-riding by creditors] may be amplified by the prevalence of complex financial instruments, such as credit derivatives, which in some cases may provide investors with incentives to hold out in the hope of forcing a default (thereby triggering a payment under the derivative contract), rather than participating in a restructuring”).

- 2 -

participants and the courts, adequately address many of the legal uncertainties arising from

sovereign restructurings. Nevertheless, from a broader stability perspective, credit derivatives

in their current form increase the likelihood of future sovereign defaults. They do this by

creating incentives for protected holders of sovereign debt to prefer a default to a voluntary

debt restructuring, since the former triggers payment from the protection seller while the latter

does not. Even when a voluntary exchange succeeds, abstention by protected creditors and

the resulting transfer or debt to protection sellers increase the likelihood of holdout litigation.

In addition, the concentrated structure of the sovereign credit derivatives market may increase

global systemic risk, and the availability of credit protection may encourage excessive lending

in emerging markets.

Moreover, these deficiencies are not the result of faulty drafting, as the current

contractual framework adequately reflects the interests of market participants. Whether the

discrepancy between these legitimate interests and the need to facilitate orderly restructurings

of unsustainable sovereign debt will in fact lead to undesirable consequences is ultimately an

empirical matter. What this discrepancy does, however, is point to the need for an improved

sovereign debt restructuring process. An analysis of the treatment of newly-adopted

collective action clauses under the credit derivatives rules reveals that they may alleviate the

incentives problems created by credit derivatives, assuming that the rules governing them are

properly interpreted by the courts.

Part I of this paper outlines the development of the secondary market in sovereign

bonds, the court decisions that have made future restructurings vulnerable to the actions of

recalcitrant creditors, and recent proposals for reform of the debt restructuring process. Part II

reviews the characteristics and usage of credit derivatives, the principal features of the

- 3 -

sovereign credit derivatives market, and the market failures potentially associated with the use

of credit protection. It also introduces the standardized documentation used in credit

derivatives markets, and the debate over the restructuring credit event. Part III analyzes the

applicability of these standard terms to the various possible forms a sovereign debt

restructuring may take, and the resulting impact on the incentives of protection sellers. It also

covers the treatment of collective action clauses and a potential sovereign bankruptcy regime

under current contractual terms. Finally, it explores the potential systemic risk and financial

stability problems caused by the structure of the sovereign credit derivatives market.

I. THE SOVEREIGN DEBT RESTRUCTURING DEBATE

A. General Background: Crises and Restructurings7

1. The 1980s Sovereign Debt Crisis

Throughout the 1970s and into the early 1980s, commercial banks in New York,

buoyed by large inflows of funds from oil-exporting countries, made substantial syndicated

loans to sovereigns and private debtors in developing countries, notably in Latin America.

These investments seemed promising, given the rising prices commanded by these countries’

exports in the world markets and the weak lending opportunities in developed countries.8

The optimism, however, did not last. In an effort to counter inflationary pressures, the

U.S. Federal Reserve sharply increased interest rates in 1981, thus substantially increasing the

7 The history of the 1980s sovereign debt crisis has been told many times. Accordingly, I will only provide a brief summary of the relevant background. For a concise account, see Philip J. Power, Note, Sovereign Debt: The Rise of the Secondary Market and Its Implications for Future Restructurings, 64 FORDHAM L. REV. 2701 (1996); see also ROSS P. BUCKLEY, EMERGING MARKETS DEBT : AN ANALYSIS OF THE SECONDARY MARKET (1999); Jessica W. Miller, Solving the Latin American Sovereign Debt Crisis, 22 U. PA. J. INT’L ECON. L. 677 (2001). 8 See Power, id. For an overview of competing explanations of the crisis, see Miles Kahler, Politics and International Debt: Explaining the Crisis, in THE POLITICS OF INTERNATIONAL DEBT 11 (Miles Kahler ed., 1986).

- 4 -

interest payments due by foreign debtors. A steep decline in export prices due to

overproduction, along with a rise in the price of their oil imports, fragilized many developing

countries’ economies and depleted their foreign currency reserves. In addition, the failure of

costly development strategies based on large-scale infrastructure and industrial projects

further impaired their capacity to service their foreign debt.9

These difficulties culminated in the Latin American debt crisis of the 1980s. In 1982,

Mexico announced that it could no longer service its foreign debt and would seek

arrangements with its creditors. Several other Latin American countries, including Argentina

and Brazil, soon followed suit.

The crisis placed the banks in a precarious situation. In many cases, their exposure to

Latin American debtors exceeded their capital.10 Thus, an outright default, which would have

required them to write off the principal amount of the loans, would have threatened their

solvency. Accordingly, they agreed to reschedule some of the debt and extend bridge loans to

allow their creditors to meet interest payments and avoid defaulting on the original loans. The

general expectation was that, in due course, the debtors’ liquidity problems would subside and

regular payments would resume. In the meantime, the continued interest payments allowed

the banks to keep the sovereign loans on their balance sheets as performing loans.

As Latin American economies failed to recover and a cycle of reschedulings ensued,

however, it became increasingly clear that they were not a long-term solution to the ongoing

crisis. The 1985 Baker Plan, which involved additional loans by both commercial banks and

9 See SEBASTIAN EDWARDS, CRISIS AND REFORM IN LATIN AMERICA: FROM DESPAIR TO HOPE 1-6 (1995); Javier Corrales, Market Reforms, in CONSTRUCTING DEMOCRATIC GOVERNANCE IN LATIN AMERICA 74 (Jorge I. Dominguez & Michael Shifter eds., 2d ed. 2003). 10 See Power, supra note 7, n. 40; HAL S. SCOTT, INTERNATIONAL FINANCE: TRANSACTIONS, POLICY AND REGULATION (11th ed. forthcoming 2004), Draft ch. 16, at 4-5 [hereinafter INTERNATIONAL FINANCE].

- 5 -

multilateral institutions in return for the implementation of IMF-monitored austerity plans,

failed to reduce the debtor countries’ existing burden and the need for further loans.11

2. The Secondary Market and the Brady Plan

During that period, a secondary market in sovereign debt started to develop. Several

banks sought to reduce their exposure to foreign sovereigns by selling off parts of their loans

at a discount, and writing off the difference as a loss. By that time, the banks had sufficiently

bolstered their capital that the gradual write-offs no longer affected their balance sheets as

dramatically as they would have at the outset of the debt crisis. Secondary demand for these

loans was fueled by the willingness of some sovereign debtors to negotiate debt-equity swaps,

under which foreign creditors could obtain full payment of the face amount of their debt,

albeit in local currency and on the condition that the proceeds by reinvested locally. 12

Although the banks themselves were often not in a position to benefit from such offers, they

attracted many other investors who saw them as an opportunity to enter these markets at a

substantial discount. 13 Alongside these investors, however, others also purchased for

speculative purposes, with a view to profiting from later appreciation of the debt as the debtor

country’s economy recovered.

As a result of this development, substantial market interest arose for sovereign debt.

In 1989, U.S. Secretary of the Treasury Nicholas Brady announced a new plan to reduce the

indebtedness of developing countries. Under the Brady Plan, commercial banks agreed to 11 On reschedulings and the Baker Plan, see Ross P. Buckley, Rescheduling as the Groundwork for Secondary Markets in Sovereign Debt, 26 DENV. J. INT’L L. & POL’Y 299 (1998). 12 See Steven M. Cohen, Comment, Give Me Equity or Give Me Debt: Avoiding a Latin American Debt Revolution, 10 U. PA. J. INT’L BUS. L. 89 (1988); Daniel H. Cole, Debt-Equity Conversions, Debt-for-Nature Swaps, and the Continuing World Debt Crisis, 30 COLUM. J. TRANSNAT’L L. 57 (1992); Derek Asiedu-Akrofi, A Comparative Analysis of Debt-Equity Swap Programs in Five Major Debtor Countries, 12 HASTINGS INT’L & COMP.L.REV. 537 (1989). 13 See Power, supra note 7, at 2716-17.

- 6 -

repackage the remaining loans into bonds offered to the public and freely tradable on

secondary markets. The proceeds of the bonds were then used to retire the loans, with the

banks taking a write-off on the original face amounts. This securitization approach allowed

the banks to exit the debt rescheduling cycle and largely eliminate their exposure to sovereign

debtors, albeit at the price of recognizing substantial losses. Public investors, for their part,

benefited from significant discounts, and from a new collateralization structure under which

principal and interest payments were secured by zero-coupon U.S. Treasury bonds bought by

the debtors with IMF support.14

The Brady Plan was widely regarded as a success, as it allowed for substantial debt

reduction by engaging the private sector while improving the financial soundness of major

bank lenders. Several countries, including Mexico, Costa Rica, Venezuela, Uruguay,

Argentina and Brazil converted their loans to Brady bonds, some of which were retired in

favor of more marketable regular eurobonds when the debtor’s finances improved.

B. The Holdout Problem

1. Origins of the Holdout Problem

The Brady Plan, despite its merits, did not solve all the problems facing sovereign

debtors. The financial situation of many remained precarious. Moreover, a consensus view

developed in international policy circles that the dispersion of sovereign debt resulting from

secondary trading and the issuance of Brady Bonds complicated the prospects for orderly debt

restructurings in future crises.

14 For a detailed examination of the Brady Plan and the evolution of the secondary market in sovereign debt, see BUCKLEY, supra note 11.

- 7 -

During the 1980s, when syndicated bank loans were the norm, negotiations between

sovereign debtors and their creditors were relatively simple. An ad hoc Bank Advisory

Committee, led by the banks with the largest exposures to the particular sovereign, would be

set up to negotiate with the distressed sovereign. Individual banks with smaller exposures

would be discouraged from holding out on the restructuring or free riding on the bridge loans

extended by others through pressure from peers, regulatory agencies and multilateral

institutions.15 Moreover, the legal remedies available to a creditor who decided to hold out

were largely untested. Thus, over that period, holdout creditors were not a substantial

concern.

The rise of the secondary market and the Brady Plan, however, led to the emergence

of a sizeable category of creditors with little or no structural incentives to agree to a

restructuring. These investors are largely immune from the pressures mentioned above, given

their lack of ongoing relationships with either the debtors or other creditors. Even though

they may have bought their debt at a substantial discount, they have little reason to settle for

anything less than the face amount if they can secure it. This made the legal regime

surrounding sovereign restructurings a matter of increasing concern.

2. Sovereign Debtors in U.S. Courts

For a long time, the prospects for effective legal enforcement of sovereign debt in the

United States were very limited. From the mid-1980s on, however, the obstacles to

15 See the United States Statement of Interest in CIBC, cited in Power, supra note 7, at 2751; INTERNATIONAL FINANCE, supra note 10, Draft ch. 16 at 5-6.

- 8 -

successfully suing sovereign debtors in U.S. courts began to fall.16 The groundwork for these

developments had been laid decades before, by a change in the international law of sovereign

immunity. In a famous letter to the Attorney General of the United States dated May 19,

1952, Jack Tate, legal adviser to the State Department, expressed the Department’s adoption

of the so-called restrictive theory of state immunity. Under this theory, “the immunity of the

sovereign is recognized with regard to sovereign or public acts (jure imperii) of a state, but

not with respect to private acts (jure gestionis).” 17 The restrictive theory, codified in

extensive detail by the Foreign Sovereign Immunities Act of 1976,18 was generally understood

as encompassing borrowing in the United States within the “commercial activities” with

respect to which foreign sovereigns are subject to the jurisdiction of U.S. courts. The

Supreme Court confirmed this in the 1992 case of Weltover v. Argentina,19 where it held that

Argentina’s postponement of payments on bonds payable in New York constituted a

commercial act of the sovereign outside the United States, and that the act in question caused

a direct effect in the United States. U.S. courts therefore had jurisdiction under Section

1605(a)(2) of the FSIA.

In any case, sovereign debt instruments normally include waivers of sovereign

immunity, which also suffice to ground U.S. jurisdiction under the FSIA20 and customary

16 See generally Samuel E. Goldman, Mavericks in the Market: The Emerging Problem of Hold-Outs in Sovereign Debt Restructuring, 5 UCLA J. INT’L L. & FOREIGN AFF. 159 (2000); Ronald J. Silverman & Mark W. Deveno, Distressed Sovereign Debt: A Creditor’s Perspective, 11 AM. BANKR. INST. L. REV. 179 (2003). 17 Letter from Jack B. Tate, Acting Legal Adviser, Department of State, to the Attorney General of the United States (May 19, 1952), in 24 DEP’T ST. BULL., 1952. 18 Foreign Sovereign Immunities Act, Pub. L. No. 94-583, 90 Stat. 2891 (codified at 28 U.S.C. §§1130, 1602-11 (2000)) [hereinafter the FSIA]. The FSIA was the first in a series of modern national codifications of modern sovereign immunity law, which integrated the restrictive immunity theory. See, e.g., the State Immunity Act, 1978, 26 Eliz. II, 33 (England), 17 I.L.M. 1123. 19 504 U.S. 607 (1992). 20 See FSIA, supra note 18, s. 1605(a)(1).

- 9 -

international law.21 Other court decisions in the 1980s dismissed defenses based on the act of

state doctrine22 and Article VIII(2)(b) of the IMF Articles of Agreement.23 Thus, in the

1990s, the enforceability of sovereign debt was well settled in principle, and the courts’

attention turned to a series of new defenses raised by sovereign debtors in an attempt to

protect the restructuring process itself.

The most important of these developments related to the holdout creditor problem.

When a sovereign debtor proposes a restructuring to its private creditors (for instance, in the

form of an offer to exchange their existing bonds for modified ones), each creditor has two

options: she can either accept or refuse the offer. In most cases, the offer will be the result of

negotiations between the sovereign and representatives of the largest creditors. Thus, these

creditors will have tentatively expressed their agreement with the restructuring terms and their

intention to participate in the exchange.

A smaller creditor, however, might see the situation differently. Suppose the

restructuring plan embodies a fair balance between the rights of creditors and the sovereign’s

ability to pay, so that the creditors as a group could not obtain better terms from the debtor.

Nevertheless, if the law provides an individual creditor with a remedy which allows her to

hold out from the restructuring and receive full payment on the debt she holds, she will have a

strong incentive to do so. Even though the debtor does not have the resources to make full

payment to all its creditors, it will probably be able to do so with respect to the small amount 21 See IAN BROWNLIE, PRINCIPLES OF PUBLIC INTERNATIONAL LAW 343-44 (5th ed. 1998); MALCOLM N. SHAW, INTERNATIONAL LAW 516-17 (4th ed. 1997). 22 See, e.g, Allied Bank International v. Banco Credito Agricola de Cartago, 757 F.2d 516 (2d Cir. 1985), cert. dismissed, 473 U.S. 934 (1985). 23 This Article renders unenforceable an “exchange contract” contrary to exchange control regulations imposed by an IMF member country. U.S. Courts, however, have interpreted this provision restrictively, holding that a loan contract which requires payment to be made in a foreign currency is not an “exchange contract.” See, e.g., Libra Bank Ltd. v. Banco National de Costa Rica, 570 F.Supp. 870 (S.D.N.Y. 1983).

- 10 -

held by the holdout. The problem, of course, is that other creditors will not look kindly upon

this unequal treatment. To avoid the risk of receiving only partial payment while the holdout

emerges unscathed, even creditors who initially agreed to the plan will be tempted to

withdraw from the offer and sue for payment. In the worst scenario, the restructuring will

collapse as creditors race to the courts to attach the debtor’s limited assets before they run out.

Three cases from the mid-1990s provide concrete illustrations of the holdout problem

and the policies elaborated by New York courts with respect to sovereign restructurings.

These cases, by bringing the holdout problem to the fore of the policy debate, were also a

major cause of the momentum enjoyed by the reform proposals that followed.

a. CIBC Bank & Trust Co. (Cayman) Ltd. v. Banco Central do Brasil

The first significant holdout litigation arose out of a series of Brazilian debt

restructurings. In 1988, following several reschedulings, an agreement was reached between

the country and a vast majority of its creditors. Under this Multi-Year Deposit Facility

Agreement (MYDFA), which covered over $60 billion of Brazil’s debt, the Central Bank of

Brazil became the obligor of the restructured debt. The MYDFA included provisions dealing

with assignments by creditors, and made acceleration upon an event of default conditional on

a vote of creditors holding at least 50% of the debt.

Then, in 1992, Brazil and its Bank Advisory Committee agreed to a further

restructuring under the Brady Plan. Creditors were initially given two options. They could

exchange the full face amount of their MYDFA debt for collateralized bonds with a lower,

fixed interest rate, or they could instead opt for uncollateralized bonds with a rising interest

rate. After many creditors committed to the offer, however, Brazil modified its terms and

- 11 -

requested that they convert at least 35% of their debt to collateralized, deep-discount bonds

with a floating interest rate. Despite this change, the vast majority of Brazil’s creditors

participated in the restructuring, with one major exception. The Darts, a wealthy Florida

family acting through a series of trusts, had been buying Brazilian debt in the secondary

market at a large discount since 1991. They now held approximately $1.4 billion in MYDFA

debt. After the Brady deal was finalized, the Darts refused to participate and insisted that

Brazil convert their entire holdings to uncollateralized bonds in accordance with the original

offer.

Brazil refused, and adopted a strategy aimed at preventing the Darts from accelerating

the MYDFA debt. Banco do Brasil, a state-owned bank that held a large amount of MYDFA

debt, was ordered partially to withdraw from the restructuring and retain about $1.6 billion of

debt so that the Darts could not trigger acceleration by a majority vote. The Darts, through

their nominee CIBC, then sued Brazil, its central bank and Banco do Brasil. They asked the

Court to issue a judgment for the full principal and interest of their MYDFA debt, on the

grounds that the defendants had breached the agreement and had acted in bad faith by

colluding with Banco do Brasil to prevent them from accelerating the loans.

In CIBC Bank & Trust Co. (Cayman) Ltd. v. Banco Central do Brasil,24 the Southern

District of New York held that the defendants’ failure to pay interest on the Darts’

unconverted MYDFA debt was a breach of contract. The Darts were thus entitled to unpaid

interest in an amount of close to $60 million. The court held, however, that the Darts could

not accelerate the principal amount, as they did not hold a majority of the debt. Preska J.

24 886 F.Supp. 1105 (S.D.N.Y. 1995).

- 12 -

rejected the Darts’ argument that an “implied covenant of good faith and fair dealing”25

prohibited Brazil from using Banco do Brasil’s holdings to obstruct acceleration. The clear

language of the MYDFA, the plaintiffs’ knowledge of the relationship between Brazil and

Banco, and other provisions suggesting that the drafters would expressly have excluded

Banco’s share had they so wished, all militated against the Darts’ claim.

CIBC is an important case for several reasons. First, although the defendants

ultimately succeeded in preventing the Darts from recovering the full principal amount of

their debt, this success was based exclusively on a contractual defense. In other words, had

Banco do Brasil not been in a position to outvote the Darts, they would in all likelihood have

been able to accelerate the debt and obtain judgment for the full principal amount. Whether

they would have had any significant practical remedy against Brazil, however, is uncertain.

Second, the United States submitted a Statement of Interest to the Court arguing that

the plaintiffs’ claim should be dismissed. This was remarkable because in a previous case,

Allied Bank,26 the government had filed a brief in support of the plaintiffs’ position that

international comity should not be extended to protect foreign exchange controls imposed by

Costa Rica. The controls prevented the defendant banks from making payments on their

foreign debt.27 In accordance with the Statement of Interest, the Second Circuit noted that

U.S. policy on sovereign debt restructurings was “grounded in the understanding that, while

parties may agree to renegotiate conditions of payment, the underlying obligations to pay

nevertheless remain valid and enforceable.”28 By contrast, in CIBC, the United States pointed

25 Id. at 1114. 26 Supra note 22. 27 See Power, supra note 7, at 2740-41. 28 757 F.2d 516, 519; see Power, id., at 2741.

- 13 -

out that the growing secondary market in sovereign debt weakened the factors that previously

limited the incentives to litigate: buyers in the secondary market did not have the same long-

term interests as the original lenders, and their expectations of full recovery should be

informed by the fact that they acquired the debt at a substantial discount. The apparent

change of policy in CIBC thus appeared to provide sovereign debtors with a powerful new

weapon in their efforts to fend off holdout creditors.

b. Pravin Banker Associates, Ltd. v. Banco Popular del Peru

The limits of this policy, however, soon became clear in Pravin Banker Associates,

Ltd. v. Banco Popular del Peru.29 Pravin had purchased $9 million of Banco Popular debt

from Mellon Bank on the secondary market in 1990. The debt was guaranteed by the

Republic of Peru. Since 1984, the debt had been in default following a national liquidity

crisis and the imposition of foreign exchange restrictions. While interest payments were

made between 1984 and 1992, the maturity date had long gone by without repayment of the

principal. In 1989, most of Peru’s external creditors, including Mellon, had filed lawsuits to

prevent the statute of limitations from expiring on the outstanding debts. Shortly thereafter,

however, the same creditors had agreed to stay their lawsuits and negotiate a Brady Plan

restructuring of Peru’s foreign debt with the new Fujimori government. The stay was

conditioned upon none of the individual lawsuits being allowed to proceed on its own.

Although Pravin resold part of the debt shortly after its acquisition in 1990, it still held

$1,425,000 in 1992. At the beginning of that year, Banco Popular stopped making interest

payments and Peru’s central bank appointed a committee of liquidators to dissolve the failing

29 109 F.3d. 850 (2d Cir. 1997).

- 14 -

bank. Pravin refused to participate in the liquidation proceedings or the Brady Plan. Instead,

it filed suit against Banco Popular and Peru for the full amount of the debt. The defendants

argued that the action should be dismissed on grounds of international comity. In their

submission, allowing the Pravin lawsuit to proceed would allow all of Peru’s other creditors

to reawaken their own suits, create a race to attach Peruvian assets, and disrupt the vital

structural reform efforts underway in Peru.

After initially granting several stays to allow the liquidation proceedings to continue

and to resolve numerous preliminary questions, the Southern District of New York granted

Pravin’s motion for summary judgment. The Second Circuit upheld the District Court’s

decision on appeal. Calabresi J. observed that the doctrine of international comity, under

which U.S. courts may recognize foreign proceedings with extraterritorial effects in the

United States, does not apply when “doing so would be contrary to the policies or prejudicial

to the interests of the United States.30” Thus, the question was whether the U.S. policies

implicated in the suit militated in favor of or against recognition of the Peruvian liquidation

proceedings. Calabresi J., citing both CIBC and Allied Bank, held that:

First, the United States encourages participation in, and advocates the success of, IMF foreign debt resolution procedures under the Brady Plan… Second, the United States has a strong interest in ensuring the enforceability of valid debts under the principles of contract law, and in particular, the continuing enforceability of foreign debts owed to United States lenders… This second interest limits the first so that, although the United States advocates negotiations to effect debt reduction and continued lending to defaulting foreign sovereigns, it maintains that creditor participation in such negotiations should be on a strictly voluntary basis. It also requires that debts remain enforceable throughout the negotiations.

30 Id. at 854.

- 15 -

He accordingly held that Pravin was entitled to enforce the debt and obtain judgment for the

full amount. Pravin, however, eventually abandoned its litigation strategy and participated in

the Peruvian exchange offer, after failing to locate Peruvian assets abroad it could attach.31

As will be discussed below, the availability of an international comity defense is

premised on U.S. policy, which may change or be reinterpreted in the future. After Pravin,

however, it seems clear that under the current interpretation, this doctrine will not normally

provide a defense against lawsuits by holdout creditors.

c. Elliott Associates, L.P. v. Banco de la Nacion

The rulings in CIBC and Pravin emboldened speculators, some of which began to

purchase discounted sovereign debt on the secondary market for the purpose of negotiating

with the debtor for full payment (or at least a substantial premium over the market price). In

Elliott Associates, L.P. v. Banco de la Nacion,32 Elliott, a hedge fund specialized in distressed

debt, had purchased working capital debt of Banco de la Nacion, a Peruvian bank, of a face

amount of $20.7 million. Here, as in Pravin, the debt was guaranteed by the Republic of Peru

and was purchased at a large discount.33 There was evidence that Elliott deliberately delayed

the purchase until the Second Circuit rendered its judgment in Pravin.34 Shortly thereafter,

Elliott closed the transaction and notified the debtors that it intended to negotiate repayment

terms. Banco de la Nacion and Peru refused to negotiate, alleging that the assignment to

31 See Mark A. Cymrot, Barricades at the IMF: Creating a Municipal Bankruptcy Model for Foreign States, 36 INT’L LAW. 1103, 1110 (2002). 32 194 F.3d 363 (2d Cir. 1999). 33 Elliott paid approximately $11.4 million. Id., 367. 34 Id.

- 16 -

Pravin was void on grounds of champerty under s. 489 of the New York Judiciary Law.35

Elliott, in turn, refused to participate in Peru’s ongoing Brady Plan restructuring and filed suit

against the debtors.

After a tortuous procedural history, the Second Circuit eventually decided in favor of

Elliott. Essentially, Section 489 prohibits taking certain debts by assignment “with the intent

and for the purpose of bringing an action or proceeding thereon.” Michel J. held that Elliott’s

primary aim in acquiring the debt was to be paid in full, by suing if necessary. Therefore,

Elliott’s intent to sue was merely incidental and was not the primary purpose of the

acquisition. Moreover, it was contingent, since the defendants could have avoided the suit by

making payment in accordance with the terms of their contract.36 This conclusion was

reinforced by the policy, enunciated in Pravin, that sovereign debt restructurings be voluntary

and that the debts remain enforceable throughout the process. Following this appeal and

remand, Elliott obtained judgment for the full principal amount of the debt in July 2000.

The case, however, was not over. Elliott faced another hurdle. As in many other

cases involving sovereigns, the prospects for successfully attaching assets were limited.

Peruvian courts would almost certainly have refused to recognize the judgment, on public

policy or similar grounds. Thus, Elliott had to find and attach Peruvian assets abroad. Most

sovereign assets in foreign countries, however, are immune from attachment or seizure under

35 This Section, in relevant part, provides that “[n]o person or co-partnership, engaged directly or indirectly in the business of collection and adjustment of claims, and no corporation or association, directly or indirectly, itself or by or through its officers, agents or employees, shall solicit, buy, or take an assignment of a bond, promissory note, bill of exchange, book debt, or other thing in action, or any claim or demand, with the intent and for the purpose of bringing an action or proceeding thereon.” 36 A champerty defense based on s. 489 had previously been raised in CIBC, supra note 24, 1110-1111. The Court, however, refused to grant the defendants’ motion to dismiss on that basis, emphasizing that champerty requires a fact-specific inquiry and that the facts indicated that other, legitimate reasons existed for transferring the debt from the Darts’ previous nominees to CIBC.

- 17 -

international law. Elliott came up with a creative solution: it would attach the payments

about to be made to Peru’s other creditors under the Brady Plan restructuring. Elliott first

obtained a restraining order against the Chase Manhattan Bank, acting as fiscal agent for Peru

in connection with the Brady bonds, but this tactic failed when Peru managed to stop the

transfer of funds to Chase in extremis. Peru then attempted to effect payments through

Euroclear, a European clearing and settlement organization based in Brussels, but Elliott

obtained a restraining order from the Brussels Court of Appeals.37 At this point, Peru opted to

settle with Elliott for $56.3 million rather than defaulting on the Brady bonds, whose grace

period was running out.38 In effect, Elliott was able to place Peru in a position where the only

choices were to pay, or to see its carefully negotiated debt restructuring collapse, along with

its efforts at economic reform.

3. Towards the Nightmare Scenario?

After Elliott, the legal techniques for preventing holdout creditors from disrupting the

sovereign debt restructuring process appeared increasingly inadequate. A virtual smorgasbord

of more or less creative defenses had almost invariably failed. Sovereign immunity and the

act of state doctrine provide little or no protection to sovereigns who choose to raise capital in

the U.S. markets.39 Restrictions on assignability in loan instruments had proved ineffective to

prevent a substantial secondary market from emerging.40 In any case, these restrictions are

increasingly uncommon in the present era of freely tradable sovereign bonds.

37 The order was granted on appeal from the Commercial Court, which had denied Elliott’s claim. See Elliott Assocs., L.P., General Docket No. 2000/QR/92 (Court of Appeals of Brussels, 8th Chamber, Sept. 26, 2000). 38 About this episode, see G. Mitu Gulati & Kenneth N. Klee, Sovereign Piracy, 56 BUS. LAW. 635, 635-36 (2001). 39 See Weltover, supra note 19; Allied Bank, supra note 22. 40 See, e.g., Elliott, supra note 32.

- 18 -

International comity41 and the IMF Articles of Agreement42 do not, in their current

interpretation by U.S. courts, protect sovereign debtors from lawsuits during a restructuring.

Although U.S. policy may change and inform a more robust approach to comity, there are

little signs of such a development at this time.43 The New York law of champerty, for its part,

no longer poses a substantial obstacle to holdout creditors. While earlier court rulings

rejecting this defense had insisted on a fact-specific inquiry, the Second Circuit holding in

Elliott leaves little doubt that, as a matter of law, a creditor who bought its debt – even at a

discount – with an intent to obtain full payment is not barred from suing the debtor under

Section 489.44

As a result, sovereign debtors are now largely dependent on defenses based on the

terms of the obligations themselves, as well as on any tactics that may be put in place with the

cooperation of the bulk of their creditors.45 One such tactic is the “exit consent,” which

involves having creditors consent, as part of an exchange agreement, to amendments to the

nonfinancial terms of the old debt instruments that make them unattractive to holdout

creditors.46 As a result, all creditors will have an incentive to take part in the exchange;

otherwise, they will be stuck with largely worthless securities. Exit consents have played a

41 See Pravin, supra note 29. 42 See Libra Bank, supra note 23. 43 But see Christopher C. Wheeler & Amir Attaran, Declawing the Vulture Funds: Rehabilitation of a Comity Defense in Sovereign Debt Litigation, 39 STAN. J. INT’L L. 253 (2003) (advocating such a reinterpretation). One source of optimism is the United States’ recent Statement of Interest supporting Argentina’s interpretation of the pari passu clause against its creditors’ position that the Elliott interpretation should prevail. See Statement of Interest of the United States, Macrotecnic Int’l Corp. v. Republic of Argentina, 02 CV 5932 (TPG), January 12, 2004. 44 See Elliott, supra note 32. 45 See Lee C. Buchheit & G. Mitu Gulati, Sovereign Bonds and the Collective Will, 51 EMORY L.J. 1317 (2002). 46 As will be seen below, amendments to financial terms require unanimity, but other terms may normally be amended by a majority or supermajority of creditors. See Lee C. Buchheit & G. Mitu Gulati, Exit Consents in Sovereign Bond Exchanges, 48 UCLA L. REV. 59 (2000); Stephen Choi & G. Mitu Gulati, Why Lawyers Need to Take a Closer Look at Exit Consents, INT’L FIN. L.R., Sept. 2003, at 15.

- 19 -

significant role in recent years, allowing sovereigns such as Ecuador and Uruguay

successfully to restructure their foreign debt. Some commentators, however, argue that exit

consents, while useful as a short-term solution, are insufficient to improve and stabilize the

restructuring process. First, it is unclear whether a determined holdout creditor would be

fazed, as nonfinancial amendments cannot at the end of the day prevent him from filing suit.

Second, whether courts would enforce such seemingly oppressive amendments against

minority debtholders appears questionable. 47 Finally, following the recent use of exit

consents, creditors have become aware of the tactic and reportedly demand better protection

in new bond issues against nonfinancial amendments.48

The Elliott Belgian restraining order caused additional consternation in financial and

policy circles. Up to that point, it had been widely assumed that the potential damage

holdouts could inflict was limited by the difficulty of finding attachable assets of the

sovereign abroad. The Belgian decision changed this by raising the fear that holdouts would

simply hold ongoing restructurings hostage in order to obtain preferential payment. The

decision thus lent added urgency to the restructuring debate. Its interpretation of the pari

passu clause has severely criticized by several scholars and practitioners, and will likely be

repudiated by courts.49 However, even if the pari passu clause were reinterpreted so as to

prevent holdout creditors from obtaining restraining orders against payments to other 47 See Hal S. Scott, A Bankruptcy Procedure for Sovereign Debtors?, 37 INT’L LAW. 103, 117 (2003) [hereinafter Bankruptcy Procedure?]. 48 See id., at 117-18. 49 See Gulati and Klee, supra note 38; Lee C. Buchheit & Jeremiah S. Pam, The Pari Passu Clause in Sovereign Debt Instruments, Working Paper, Harvard Law School, Program on International Financial Systems (Draft of December 11, 2003), available at www.law.harvard.edu/programs/pifs/pdfs/buchheitpam.pdf; Institute of International Finance, Letter from Charles Dallara, Managing Director, to Chancellor Gordon Wood, April 9, 2002; but see William W. Bratton, Pari Passu and a Distressed Sovereign’s Rational Choices (Draft of Feb. 15, 2004) (arguing that the Elliott interpretation of the pari passu clause benefits all creditors, not just holdouts, by diminishing the likelihood of default, preventing discriminatory treatment of particular classes, and improving the creditors’ negotiating position).

- 20 -

creditors, the problem might not be solved entirely. Elliott revealed the vulnerability of

restructuring payments, but its lesson goes beyond this: all foreign assets and payments by a

sovereign, including payments on other debt or for imports, or bank and security custody

accounts, could conceivably be attached, regardless of whether their use had anything to do

with paying foreign creditors (and thus regardless of the pari passu clause’s interpretation).50

While such tactics have not succeeded so far, the possibility that a creative creditor

might find a viable technique and open a path for others, as Elliott did, cannot be overlooked.

In both Pravin and Elliott, the sovereigns were eventually able to settle and the restructurings

went ahead. The sums at play, however substantial, were clearly not beyond their ability to

pay – $1.4 million in face amount and $56.3 million in settlement, respectively. But what if a

creditor such as the Darts, who at the time of CIBC were the single largest foreign holder of

Brazilian debt with more than $1.4 billion in holdings, obtained such a judgment and

restraining order? And what if more and more foreign creditors faced with a restructuring

choose to go down the legal path Elliott has obligingly cleared for them?

The sum of all these concerns, from the official sector perspective, was that even a

single creditor now had considerable power to derail a restructuring – and thus considerable

leverage to obtain preferential payment. According to this picture, the holdout problem has

many features of the classical prisoner’s dilemma:51 each creditor involved would be better

off if all participated in the restructuring – this is the Pareto optimal outcome. Individual

creditors, however, have incentives to hold out in the hope of achieving their preferred result:

50 See Bankruptcy Procedure?, supra note 47, at 116-17. 51 It is unnecessary further to elaborate this theoretical point for the purposes of this paper. See generally DOUGLAS G. BAIRD, ROBERT H. GERTNER AND RANDAL C. PICKER, GAME THEORY AND THE LAW (1994) for the applicability and limits of game-theoretical models to complex legal problems.

- 21 -

obtaining full payment while others have to cope with the restructuring terms. The actions of

creditors trying to achieve this, however, lead to the worst possible outcome – the collapse of

the restructuring and potential default, with adverse consequences for all creditors, the

sovereign, and its people.52 In other words, on that view, while the best interest of the

creditors is to agree unanimously to a restructuring, this outcome is prevented by a free rider

problem that can only be solved by the adoption of some form of collective action

mechanism. This concern dovetailed with the questions raised by the series of financial crises

that shook emerging markets from 1997 on, and led to a series of proposals to reform the

sovereign debt restructuring process.

4. The 1990s Currency Crises and the Role of the IMF

In addition to the restructuring problems caused by the rise of holdout creditors, in the

mid- to late 1990s a series of financial disasters shook the world’s confidence in the G-7 and

the IMF’s crisis-handling methods. When a steep devaluation of the Mexican peso increased

the cost of servicing the country’s dollar-indexed notes in 1994-95, the United States and the

IMF put together an enormous rescue package totaling more than $49 billion. The package

succeeded in stabilizing the peso and the loans were eventually repaid.53

The East Asian crisis of 1997 did not, for the most part, involve restructuring widely-

traded sovereign debt instruments.54 The foreign exposure of the largest debtor, South Korea,

52 See, e.g., Adam Thomson, Argentine Bond Default Hits its Pensioners Hardest, FIN. TIMES, March 13-14, 2004, at 2. 53 For an account of the Mexican crisis, see John H. Chun, Note, "Post-Modern" Sovereign Debt Crisis: Did Mexico Need an International Bankruptcy Forum?, 64 FORDHAM L. REV. 2647 (1996). 54 For a detailed journalistic account of the East Asian crisis and its repercussions in individual countries, see PAUL BLUSTEIN, THE CHASTENING (2003).

- 22 -

consisted primarily of short-term dollar loans extended to Korean banks by foreign banks.55

While these loans had previously been rolled over at maturity, the bursting of a real estate

bubble in Thailand contributed to a market-wide loss of confidence in East Asian economies.

By mid-1997, foreign banks were refusing to roll over their loans. Although this was not

strictly speaking a sovereign debt crisis, the Korean government would not allow the banks to

fail. It did not, however, have the foreign reserves to support them, and it turned to the IMF

for support as its reserves dwindled and its currency came under intense devaluation pressure.

In December of 1997 and January of 1998, the IMF, along with the World Bank, the

Asian Development Bank and the G-7 countries, committed enormous standby loans to South

Korea. The loans, however, did not stem the outflow and, although South Korea avoided a

default, at least $8 billion in IMF funds was in effect paid out to the foreign banks that called

their loans.56 When the crisis spread to Russia, however, it was not able to support the

outflows caused by its short-term treasury bills and defaulted on its debt in August 1998.

After this default, Russia was largely free unilaterally to restructure its obligations, but at the

price of effectively shutting down its access to foreign capital and causing substantial

domestic economic hardship.57

With hindsight, however, the Russian default did not indicate that the IMF had

become unwilling to mount large rescue packages in an effort to stem financial crises. It did

so again when Turkey faced a currency crisis in 2001, as foreign investors sold their lira-

denominated government securities en masse and repatriated the proceeds. 58 Likewise,

55 See Bankruptcy Procedure?, supra note 47, at 107. 56 See id. 57 See id., at 108. 58 See id., at 109-110.

- 23 -

although Argentina eventually defaulted in December 2001 after the IMF refused to extend

further loans because the government had not complied with budgetary conditions,

disbursement of funds eventually resumed and the current Argentine government has

negotiated increasingly favorable terms with the IMF.59 Throughout 2002, the IMF also

provided massive loans to allow Brazil to whether large capital outflows stemming from

political and economic uncertainty.60

Throughout this period, the IMF’s approach was strongly criticized. Many

commentators argued that effectively bailing out foreign creditors created moral hazard by

encouraging reckless lending to emerging markets. On this view, the taxpayer-funded IMF

loans largely went to make whole sophisticated investors who were fully aware of the risks

associated with the high returns they obtained in emerging markets. Others argued that IMF-

imposed fiscal and monetary austerity was an affront to economic sovereignty and actually

hindered growth instead of restoring the affected countries’ solvency. 61 This general

dissatisfaction with IMF bailouts contributed to a perceived need to reform the way in which

international debt and currency crises are handled.62

The fundamental idea is that foreign creditors should be bailed in, instead of bailed

out.63 In other words, they should be made to share the sacrifices imposed by the country’s

economic difficulties by agreeing to a reduction or other concessions in connection with the

59 See Nestor Kirchner's nimble cookery – Argentina’s default, and its deal with the IMF, ECONOMIST, Sept. 13, 2003. 60 See Bankruptcy Procedure?, supra note 47, at 110-111. 61 See, e.g., JOSEPH STIGLITZ, GLOBALIZATION AND ITS DISCONTENTS (2002). 62 For a broad-ranging examination of this problem, see BARRY EICHENGREEN, FINANCIAL CRISES AND WHAT TO DO ABOUT THEM (2003) [hereinafter FINANCIAL CRISES]. 63 See Barry Eichengreen, Bailing In the Private Section: Burden Sharing in International Financial Crisis Management, 23 FLETCHER F. WORLD AFF. 57 (1999); Group of Seven, Strenghtening the International Financial System and the Multilateral Development Banks, Report of G-7 Finance Ministers and Central Bank Governors, available at http://www.g8.utoronto.ca/finance/fm010707.htm (July 7, 2001).

- 24 -

debt they hold. This, of course, is but another word for debt restructuring, but with the greater

sense of urgency born of the financial crises of the past decade. In this light, the increasing

legal difficulties associated with voluntary restructurings appear more and more problematic.

The challenge, then, is to elaborate a mechanism that would facilitate burden sharing without

the adverse consequences that make countries and international financial institutions reluctant

to consider outright default.64

C. Reform Initiatives

Two major reform proposals emerged.65 The first is the inclusion in sovereign bonds

of so-called “collective action clauses” (CACs), which would allow a supermajority of

creditors to amend the instrument’s payment terms and other essential provisions. The second

involves the creation of an international bankruptcy mechanism, most likely under the

auspices of the IMF, to oversee and coordinate the sovereign debt restructuring process while

providing a legal framework for the determination of the creditors’ and debtor’s rights.

1. A Market-Based Solution: Collective Action Clauses

Contrary to sovereign bonds issued in London, those issued under New York law do

not allow a majority of creditors to amend the payment terms.66 While the relevant provisions

normally contemplate amendments to most non-financial terms by a majority vote,

amendments that would affect the debtor’s obligation to make timely payments require

unanimity. This approach has been traced to the Trust Indenture Act of 1939, which

prohibited CACs in corporate bonds on the basis that they were susceptible to abuse to the 64 These consequences include, in addition to the risk of disruptive lawsuits, that of contagion of the crisis to other countries. See FINANCIAL CRISES, supra note 62, at 70-71. 65 For a review of these and other options not considered here, see FINANCIAL CRISES 71-98; Peter B. Kenen, The International Financial Architecture: Old Issues and New Initiatives, 5 INT’L FIN. 23 (2002). 66 See Buchheit & Gulati, Exit Consents, supra note 46.

- 25 -

detriment of minority bondholders.67 The Act, however, does not apply to sovereign bonds,

so that there are no apparent legal obstacles to including CACs in such bonds.68

At first glance, CACs are an attractive solution to the holdout problem. If the

sovereign successfully negotiates a restructuring agreement with a majority of bondholders,

the terms of the agreement can be incorporated into the original bonds by amendment. As a

result, the restructuring terms will bind all bondholders, including those who refused to

negotiate or to agree to the restructuring. Commentators, however, have pointed out

numerous difficulties inherent to CACs. Emerging market sovereigns are reluctant to adopt

them, as the market may read the clauses as signaling the probability of future financial

difficulties. Creditors also resist CACs with low percentage requirements, out of concern that

they may be abused by the sovereign or other creditors.69 CACs with higher percentage

requirements, however, increase the risk that a holdout creditor might accumulate a sufficient

proportion of bonds – at a substantial discount – to block a restructuring. In addition,

although CACs may facilitate restructurings in respect of newly-issued bonds, they will

obviously not apply to older bonds which do not include them. This is a significant issue, as

many sovereign bonds are long-term ones, with maturities of at least five years, and often

much longer.70 Finally, even with CACs in place, simultaneous restructurings of several

67 Trust Indenture Act of 1939, 15 U.S.C. ss. 77aaa, et seq. See generally Mark J. Roe, The Voting Prohibition in Bond Workouts, 97 YALE L.J. 232 (1987). See also Bankruptcy Procedure?, supra note 47, at 110. 68 See Buchheit & Gulati, Collective Will, supra note 45. 69 For instance by coercing domestic holders to consent to a restructuring. See Bankruptcy Procedure?, supra note 47, at 121-22; compare CIBC, supra note 24. Creditors have also resisted sharing clauses, which would greatly reduce the benefits of holding out by forcing a creditor who succeeds in securing preferential payment to share it pro rate with the other creditors. See Buchheit & Pam, supra note 49 at 12. 70 See, however, Anna Gelpern, How Collective Action Is Changing Sovereign Debt, INT’L FIN. L.R., May 2003, 19 at 20 (arguing that long-term bonds are the exception, and that there is “reasonably frequent turnover” is sovereign debt stock).

- 26 -

categories of bonds, and across bonds and other kinds of debt, may remain difficult or

impossible.71

These difficulties, however, have not discouraged some sovereigns from including

CACs in recent bond issues, with some pressure from the U.S. Treasury, the IMF and the G-

7.72 In a speech delivered in April 2002, John B. Taylor, U.S. Undersecretary of Treasury for

International Affairs, urged the inclusion of contractual restructuring terms – including

collective action clauses – in future sovereign debt instruments.73 Thereafter, a February 2003

Mexican notes offering included provisions by which a majority of 75% of holders can

change the amounts payable on the notes or the applicable due dates, the payment currency,

the governing law or jurisdiction clauses, or the amendment clause itself.74 A South African

offering from May 2003 also includes similar provisions.75

71 See Bankruptcy Procedure?, supra note 47; A NEW APPROACH, supra note 6. 72 See John B. Taylor, Sovereign Debt Restructuring: A U.S. Perspective (Institute for International Economics, Washington, D.C.) (April 2, 2002); Group of Seven, supra note 63; Group of Ten, The Resolution Of Sovereign Liquidity Crises: A report to the Ministers and Governors prepared under the auspices of the Deputies, available at http://www.bis.org/publ/gten03.htm (May 1996); GROUP OF TEN, REPORT OF THE G-10 WORKING GROUP ON CONTRACTUAL CLAUSES (2002), available at www.oecd.org/dataoecd/62/51/2501714.pdf; Michael M. Phillips, Support Builds for Plan to Ease Debt Loads of Developing Nations, WALL ST. J., Sept. 17, 2002, at A16. A leading economist in this area argues that CACs strike the right balance between creditor rights and the need to facilitate restructurings, whereas a more structured mechanism would overly weaken the former. See, e.g., FINANCIAL CRISES, supra note 62, at 10-11; Barry Eichengreen & Christof Rühl, The Bail-In Problem: Systematic Goals, Ad Hoc Means, NBER Working Paper 7653 (April 2000), available at http://www.nber.org/papers/w7653. 73 See Taylor, id. 74 See, e.g., United Mexican States, 6.625% Global Notes Due 2015, Prospectus Supplement Dated February 26, 2003. The notes also provide that other provisions may be amended by a 2/3 majority. See John Authors, Mexico Pioneers a Plan to Ease Debt, FIN. TIMES, Feb. 25, 2003; John Authors, Mexico Sends Signal with Bond Clauses, FIN. TIMES, Feb. 27, 2003, at 23. 75 See, e.g., Republic of South Africa, 5.25% Notes Due May 16, 2013, Prospectus Supplement Dated May 9, 2003.

- 27 -

In April 2003, Uruguay launched an exchange offer for its outstanding debt

instruments.76 Under this complex offer, debtholders were entitled to receive new bonds

denominated in various currencies and with different maturities and interest rates, depending

on the nature of the bonds they agreed to exchange. Thus, the offer replaced the existing debt

instruments with several series of new ones. The CAC is tailored accordingly: amendments

to the payment terms of a series require a 75% majority of the holders of that series; however,

if the amendments affect more than one series, they may be effected by a vote of 85% of the

aggregate amount of such series and 66 2/3% of each series. This aggregation provision

makes it more difficult for the holders of a small series to block a larger restructuring, and

thus alleviates the difficulty noted above with respect to amendments across different

instruments. Moreover, since the exchange offer covers almost all of Uruguay’s existing

private debt, it also avoids the transition issues that were thought to threaten the viability of

collective action clauses. The Uruguay CAC is also more complete than that used in Mexico,

as it includes more demanding rules for non-reserve amendments, as well as limitations on

future exit consents77 and on issuance of new debt to dilute the existing creditors’ votes.

More recently, other countries, including Canada, Turkey, Belize, Guatemala, Panama,

Venezuela and South Korea have begun experiments with CACs.78 However, an April 2003

76 See, e.g., República Oriental del Uruguay, Exchange Offer, Prospectus Supplement Dated April 10, 2003; see also Gelpern, supra note 70; Alan Beattie, Uruguay provides test case for merits of voluntary debt exchange, FIN. TIMES, April 23, 2003, at 3. 77 See Felix Salmon, Uruguay Closes the Loop, EUROMONEY, May 2003. 78 See William W. Bratton & G. Mitu Gulati, Sovereign Debt Reform and the Best Interest of Creditors, 57 VAND. L. REV. (forthcoming 2004) (manuscript at 29, fn. 86), available at http://ssrn.com/abstract=387880. For a more exhaustive progress report, see International Monetary Fund, Progress Report to the International Monetary and Financial Committee on Crisis Resolution (Sept. 5, 2003), available at http://www.imf.org/external/np/pdr/cr/2003/eng/090503.pdf; see also Sergio J. Galvis & Angel L. Saad, Collective Action Clauses: Recent Progress and Challenges Ahead (Draft of Feb. 20, 2004).

- 28 -

Brazilian bond issuance, which includes a 85% threshold for collective action, has raised

concerns that market practice may already be fragmenting.79

Finally, at least one study of CACs suggests that they may already be more

widespread than thought by most market participants. Gugiatti and Richards claim that many

U.S. sovereign debt instruments, generally those prepared by the London offices of New York

firms, contain CACs based on British models.80 However that may be, it is clear that, as a

result of the developments noted above, practice has now outrun the institutional and

academic debate. CACs have become a reality. Whether they will ever be used successfully,

however, remains to be seen.

2. An Institutional Solution: An International Bankruptcy Court?

In the aftermath of the financial crises of the late 1990s, and in light of the limitations

of CACs discussed above, several commentators have proposed a more ambitious solution.

They would set up a mechanism, inspired by Chapter 11 of the U.S. Bankruptcy Code, which

would allow an insolvent sovereign debtor and its creditors to negotiate a debt restructuring

while being protected from individual legal proceedings. Although this idea has been floating

79 See, e.g., Federative Republic of Brazil, 10% Global Bonds Due 2007, Prospectus Supplement Dated April 29, 2003. See Felix Salmon, Brazil Goes Off on a CAC Tangent, EUROMONEY, June 2003. 80 See Mark Gugiatti & Anthony Richards, The Use of Collective Action Clauses in New York Law Bonds of Sovereign Borrowers (July 11, 2003), available at: http://ssrn.com/abstract=443840.

- 29 -

for a long time,81 it gained considerable momentum following an influential 1995 speech by

Jeffrey Sachs82 and the financial crises of the late 1990s.

So far the most credible proposal has been the one championed by Anne Krueger, the

IMF’s First Deputy Managing Director. 83 Dr. Krueger’s proposed Sovereign Debt

Restructuring Mechanism would be triggered by the debtor country, rather than imposed by

the creditors. The debtor’s choice, however, would be subject to review or approval by a third

party, such as the IMF or an arbitration tribunal, or to a majority vote of the creditors.84

Following initiation of the mechanism, litigation by creditors would be temporarily stayed

during negotiations.85 The SDRM would facilitate the provision of new funds by private

creditors during the stay by granting such new financing priority over preexisting private

indebtedness. This feature would be similar to debtor-in-possession financing under Chapter

11.

Once a restructuring agreement is reached, an affirmative vote of a qualified majority

of creditors would bind the dissenting minority to its terms, thus circumventing the holdout

problem.86 Like under Chapter 11, the voting procedures and required majority would apply

regardless of the contractual terms contained in the debt instruments themselves. Moreover,

81 The proposals for sovereign bankruptcy regimes made over the years are simply too numerous to list. For a review of the pre-1995 literature, see Kenneth Rogoff & Jeromin Zettelmeyer, Early Ideas on Sovereign Bankruptcy Reorganization: A Survey, IMF Working Paper 02/57. More recent contributions include Steven L. Schwarcz, Sovereign Debt Restructuring: A Bankruptcy Reorganization Approach, 85 CORNELL L. REV. 956 (2000); Bankruptcy Procedure?, supra note 47; Michelle J. White, Sovereigns in Distress: Do They Need Bankruptcy? (2002) BROOKINGS PAPERS ECON. ACTIVITY 287. 82 Sachs, Jeffrey D., Do We Need an International Lender of Last Resort?, Frank D. Graham Lecture, Princeton University 8 (Apr. 20, 1995), available at: www.earthinstitute.columbia. edu/about/director/pubs/intllr.pdf. 83 See A NEW APPROACH, supra note 6; Anne Krueger, New Approaches to Sovereign Debt Restructuring: An Update on Our Thinking, Address before the Institute of International Economics (April 1, 2002), available at http://www.imf.org/external/np/speeches/2002/040102.htm. 84 A NEW APPROACH, id. at 23-28. 85 Id. at 15-16. 86 Id. at 14-15.

- 30 -

unlike CACs, the SDRM could cover a broad variety of claims, and provide for a vote based

on the aggregate of all private debt of the sovereign.87 It could also potentially cover some

forms of public debt, such as bilateral debt (currently renegotiated under the Paris Club rules)

and obligations to multilateral institutions such as the IMF and the World Bank.88

The Krueger proposal attracted strong opposition from creditor groups and emerging

market sovereigns alike.89 In addition, its implementation would be impossible without U.S.

consent, but John Taylor’s April 2002 speech promoting CACs as an alternative revealed the

Treasury’s lack of enthusiasm for a formal restructuring mechanism.90 As a consequence, the

SDRM project appears to have been abandoned for the time being.91 Other proposals do not

seem to have attracted significant interest outside academic and advocacy circles. It may be,

however, that the limitations of CACs become sufficiently pressing that proposals for the

IMF’s SDRM or a similar multilateral mechanism are revived in the future.

3. Another View: The Case for Creditors’ Rights

More recently, the consensus view of the sovereign debt restructuring problem has

been challenged by commentators writing from the perspective of creditors’ rights. Several

observations are at the heart of this argument.

87 Id. at 15. 88 Id. at 17-19. 89 See, e.g., Economics Focus, A Better Way to Go Bust, ECONOMIST, Feb. 1, 2003, at 64 (“However most financiers, whether bankers or bondholders, loathe the SDRM … Fearing higher interest rates and scarcer access to capital, many emerging-market governments have also criticised the plan.”); Alan Beattie, Financial Grouping Wants New Debt Rules, FIN. TIMES, June 12, 2002; SDRM Finds Few Friends in the Markets, EUROMONEY, Nov. 2002; Galvis, Sergio J., Sovereign Debt Restructurings – The Market Knows Best, 6 INT’L FIN. 145 (2003). 90 Supra note 72. 91 Indeed, an October 2003 speech by Dr. Krueger barely mentioned the SDRM proposal, focusing instead on the inclusion of CACs in recent bond issues and the possibility of aggregating different series of claims through contractual clauses, as was done in Uruguay. See Anne O. Krueger, Market Discipline and Public Policy: The Role of the IMF (October 31, 2003). See also SDRM Is Dead, and That’s Official, EUROMONEY, May 2003 – although that rumor may in time turn out to have been exaggerated.

- 31 -

First, while the collective action problem caused by traditional unanimous amendment

clauses is real, it is not the only dynamic at play in a restructuring. The dominant theory of

why sovereigns actually pay their debt in the absence of significant legal coercion holds that

the main benefit of doing so is to maintain the sovereign’s access to international capital.92 If,

at any time, the cost of continuing to pay exceeds this benefit, the sovereign will default.

Conversely, a sovereign who has already defaulted and is thus barred from the markets, will

negotiate a restructuring and resume payments if the benefit of renewed access exceeds the

costs of doing so. In such negotiations, the relative bargaining power of the sovereign and its

creditors determines how much of the surplus will be allocated to each.

On that view, the unanimous amendment clause is one of the few advantages the

creditors have on an otherwise uneven playing field. In practice, because of the unanimity

requirement, the debtor has to meet the reservation price of a vast majority of its creditors in

order for its restructuring to succeed. Thus, creditors might rationally require such clauses, as

under a majority or supermajority restructuring procedure, their bargaining power would be

impaired.93 Moreover, non-unanimous procedures leave minority creditors vulnerable to

abuse by a majority whose interests diverge from theirs, or who collude with the debtor for

side payments, further lending or other favorable terms. For this reason, Professors Gulati

92 See Bratton & Gulati, supra note 78, at 13-17; one of the leading papers is Herschel I. Grossman & John B. Van Huyck, Sovereign Debt as a Contingent Claim: Excusable Default, Repudiation and Reputation, 78 AM. ECON. REV. 1088 (1988). Under the principal competing theory of sovereign debt, repayment cannot be supported solely by the reputation incentive, but depends on the availability of direct sanctions to creditors: see Jeremy Bulow & Kenneth Rogoff, Sovereign Debt: Is to Forgive to Forget?, 79 AM. ECON. REV. 43 (1989). For a historical case study evaluating the conclusions of competing sovereign debt theories, see James Conklin, The Theory of Sovereign Debt and Spain under Philip II, 106 J. POL. ECON. 483 (1998). As Bratton and Gulati point out, although the two theories differ considerably in the importance they attribute to legal enforcement mechanisms, both seem to support the conclusion that reducing the collective action problem among creditors reduces their share of the surplus in a restructuring. 93 See Bratton & Gulati, id.

- 32 -

and Bratton argue that the inclusion of CACs should at least be accompanied by the

recognition of a duty of good faith in intercreditor relations.94

Second, the implications of the holdout cases outlined above, including Elliott, may

have been exaggerated. No recent sovereign restructuring has collapsed because of the

actions of holdout creditors.95 On the contrary, countries such as Ecuador, the Ukraine,

Pakistan and Uruguay have successfully conducted exchange offers in recent years by using

exit consents. Some of these restructurings involved significant sacrifices on the part of their

creditors, and had considerable coercive elements.

Third, the high profile gained by so-called ‘vulture funds’ such as Elliott, who

specialize in the holdout tactics outlined above, obscures the fact that sovereign bondholders

also include large numbers of pensioners and other retail investors. 96 From an equity

perspective, such creditors should not be deprived of their limited bargaining power against

sovereigns in return from uncertain improvements in financial stability, especially since

experience reveals that the current arrangements are quite sufficient to allow orderly

restructurings.

Given the vulnerability of the current sovereign debt restructuring process to legal

action by holdout creditors, as well as the legal uncertainties surrounding the future of the

international financial architecture, the effects of the growing market for sovereign credit

94 See id. at 47; see also Andrei Schleifer, Will the Sovereign Debt Market Survive?, Harvard University, Department of Economics Working Paper (March 2003) (arguing that a collective action mechanism can only work if counterbalanced by features such as the presence of a court bound to find solutions in the best interests of creditors). 95 See Robert B. Gray, Chairman, International Primary Market Association, Remarks, 2003 AM. SOC’Y INT’L L. PROC. 223, at 226. 96 See id., at 20 (referring to “cookie jar” offerings of Argentine bonds in Germany and Japan in the 1990s); see also Richard Lapper, Creditors Unite to Seek Better Deal on Argentine Debt, FIN. TIMES, Dec. 9, 2003 (some 500,000 small investors in Europe hold defaulted Argentine bonds).

- 33 -

derivatives on the incentives faced by creditors may have a crucial impact on financial

stability. This appraisal, however, need be informed by the realization that some of the

stability concerns may have been overplayed, and that the interests of creditors ought to

remain a central consideration. I now turn to an examination of credit derivatives.

II. CREDIT DERIVATIVES

A. Credit Default Swaps

The past two decades have seen another extremely significant innovation in

international finance: the rise of derivative instruments as a major financial market.

Derivatives, as their name suggests, are financial instruments whose value and other

characteristics derive from those of an underlying instrument, index, or other variable.97 The

most basic derivatives are options, forwards and swaps – the so-called “building blocks” of

the derivatives markets.98

A stock option, for instance, is a right (but not an obligation) to buy or sell stock at a

predetermined price – the “strike price” – at a certain point in the future – the “strike date.”

The value of the option depends on fluctuations in the market price of the underlying stock.

For example, suppose A owns stock in XYZ Company, but is worried that the current market

price of the stock ($100) may go down in the future. A, however, does not wish to sell the

97 See, e.g, ALASTAIR HUDSON, THE LAW ON FINANCIAL DERIVATIVES 12 (3rd ed. 2002) (“A derivative product is a financial product the value of which is derived from another financial product.”) [emphasis in original]; INTERNATIONAL FINANCE, supra note 10, Draft ch. 14, p. 1 (“Derivatives are financial instruments whose value is based on or derived from other assets or variables”). 98 See Norman Menachem Feder, Deconstructing Over-the-Counter Derivatives, 2002 COLUM. BUS. L. REV. 677 at 691 (“[A]n option is the right to buy or sell something in the future, a forward is the obligation to buy or sell something in the future, and a swap is an exchange of periodic payment obligations in the future.”) On derivatives generally, the reference work is SCHUYLER K. HENDERSON, HENDERSON ON DERIVATIVES (2003).

- 34 -

stock outright.99 Instead, she buys a put option from B – a promise from B to buy the stock

on the maturity date if A so requests – at a strike price of $90. In exchange for that promise,

A must pay a premium up front.100

On the strike date, two things can happen. On the one hand, if the market price of the

stock is above the strike price of $90, A will choose not to exercise the put option, since she

can get a better price for her stock by selling it in the market. Thus, the option will expire

worthless, and B will keep the premium. On the other hand, if the stock price has fallen

below $90 – say, to $80 – A will exercise the put option and require B to buy the stock at the

strike price of $90. Thus, B will suffer a loss of $10. In this example, the option is

physically-settled, which means that the underlying stock is actually delivered by A to B upon

exercise. A could also buy a cash-settled option, which means that the underlying instrument

will not be delivered. Rather, B will make a payment to A for the difference between the

market value of the stock on the strike date and the strike price – here, $10. Cash-settled

options and other derivatives are common, and allow derivatives to be traded regardless of

whether one of the parties actually owns the underlying instrument.

The simple example of A’s option illustrates the main purpose for which derivatives

are used: to reallocate risk. Before buying the put option, A bore the full risk of fluctuations

in the price of her stock. If the price fell to $30, she would lose $70. After buying the option,

however, A’s downside risk is limited. The least she can get for her stock on the strike date is

$90, since this is the price for which B promised to buy the stock at A’s request. Therefore, A

99 There are countless possible reasons for this. A may not want to forfeit the possibility of a profit from a future increase in the stock price. Alternatively, A may be unable to sell the stock for regulatory or tax reasons, for instance because the stock is restricted and may not legally be resold until a certain period expires. 100 For a more exhaustive description of option types and possible combinations, see Feder, supra note 98 at 692-98; INTERNATIONAL FINANCE, supra note 10, Draft ch. 14, at 3-7.

- 35 -

has transferred the risk that the price of XYZ stock may fall below $90 to B. B has accepted

that risk, in exchange for a premium. A has retained both the risk that the price may fall by

less than $10, and the potential benefit of a price increase.

While options are relatively simple derivatives, they can be combined among

themselves and with other basic derivatives in very complex ways. Derivatives available in

the market can protect buyers from risk arising from interest rate and currency exchange rate

fluctuations, variations in the prices of securities, commodities or indexes thereof, even

against bad weather – the possibilities are virtually endless.101 While many standardized

derivatives are now traded on specialized exchanges, more complex derivatives tend to be

individually negotiated between sophisticated parties. The latter type are referred to as “over-

the-counter” (OTC) derivatives.102

The option described above was used to transfer a specific risk from A to B: the risk of

a decrease in the price of XYZ stock. This is a type of market risk, which may more generally

be defined as “the exposure to the possibility of market movements.”103 Another type of risk

typically faced by holders of debt instruments is credit risk – the risk that a debtor will default

on its obligations by reason of a deterioration in its creditworthiness.104 Credit derivatives are

designed to transfer credit risk between counterparties without transferring ownership of the

underlying asset. The simplest example of a credit derivative is the single-name credit default

swap (CDS).

101 See ROBERT J. SCHILLER, THE NEW FINANCIAL ORDER: RISK IN THE TWENTY-FIRST CENTURY (2003). 102 See Feder, supra note 98, at 678. 103 Id. at 688. 104 Cf. Id. at 706 (“Credit risk is the possibility that an obligor will fail, and, thus, be unable to meet its obligation to make a payment.”)

- 36 -

Under a CDS, a party – the “protection seller” – promises, upon the occurrence of a

“credit event” affecting the reference entity, to make a payment to the other party – the

“protection buyer.”105 The applicable credit events are negotiated between the parties and, as

will be seen below, vary from swap to swap, but they typically include events such as a failure

by the reference entity to make a payment on its obligations, and bankruptcy. Like other

derivatives, CDS can be physically-settled or cash-settled. Upon a credit event under a

physically-settled swap, the protection buyer will deliver the covered obligations of the

reference entity it owns to the protection seller, who will pay a pre-agreed amount – the “par

value” of the obligations. By contrast, under a cash-settled CDS, no delivery will take place.

Instead, the protection seller will pay the difference between the par value of the obligations

and their market value following the credit event. In return for this protection, the protection

buyer will pay a regular premium to the protection seller over the life of the swap.

For instance, suppose that A holds a $100 bond issued by JKL Company, and is

worried that JKL might go bankrupt and default on the bond in the future. To protect herself

against that risk, A buys a CDS from B, with JKL as the reference entity, bankruptcy as a

credit event, and a par value of $100 for the bond. A makes regular premium payments to B.

Later, but still during the life of the swap, JKL goes bankrupt and the value of the bond

plummets to $10. As a result of this credit event, A is entitled to a payment from B. The

nature of this payment depends on whether the CDS provides for physical settlement or cash

settlement. In the former case, A is entitled to deliver the bond to B, who will pay A the

bond’s agreed par value of $100. In the latter case, A will keep the bond, and B will make a

105 Cf. Packer & Suthiphongchai, supra note 5, at 80 (“Credit default swaps are credit protection contracts whereby one party agrees, in exchange for a periodic premium, to make a contingent payment in the case of a defined credit event.”)

- 37 -

net payment of $90 to A – the difference between the par value of $100 and the fair market

value of the bond after JKL’s bankruptcy. Note that, under a cash-settled CDS, there is no

requirement that A actually own the bond.

As can be seen, the CDS isolates credit risk from market risk. If the market price of

the bond merely decreases – for instance because interest rates have increased, or because the

market considers than JKL has become a riskier issuer – no credit event will occur and the B

will not have to compensate A under the swap. Only if a credit event occurs with respect to

JKL is the CDS triggered. Credit events are generally meant to reflect significant

deteriorations in the creditworthiness of the reference entity rather than market fluctuations, in

accordance with the general definition of credit risk mentioned above.

Although credit default swaps share important characteristics with insurance contracts,

they are not the same. First, unlike insurance contracts, which require that the beneficiary

have a material insurable interest which will be affected by the event insured against, no such

requirement applies to CDSs.106 As explained above, a party could buy protection under a

cash-settled CDS and receive a net payment upon the occurrence of a credit event, without

having to prove that she had any interest in the reference entity’s creditworthiness before

entering the swap. Even under a physically-settled CDS, the protection buyer can simply

acquire the underlying obligations after the swap has been triggered but prior to delivery.

Second, payments under insurance contracts are normally limited to the extent of the

insured’s financial loss arising from the event. Under a CDS, however, the amount of the

payment is determined exclusively by the previously agreed par value of the reference

106 See Don Bendernagel, James Hill & Brian Rance, Credit Derivatives: Usage, Practice and Issues, in SWAPS AND OTHER DERIVATIVES IN 2001 755 (Edward J. Rosen ed., 2001).

- 38 -

obligation, as well as by its market value in the case of a cash-settled CDS. This simplicity

allows claims under CDSs to be processed quickly and with a minimal probability of disputes,

unlike insurance claims, which depend on the insurer’s evaluation of the loss and often lead to

lengthy disputes and litigation. As a result of these differences, credit derivatives dealers are

not subject to insurance regulation or to common law insurance rules.107

B. The Market for Credit Derivatives

1. Usage of Credit Derivatives

The initial impetus for the development of credit risk transfer activities was the desire

of banks better to manage their risk exposures. From the 1970s on, banks began to use

techniques such as syndication and securitization to fund residential mortgage portfolios. The

attractiveness of such techniques increased following the 1982 sovereign debt crisis, as the

capital of several large international banks dwindled. The adoption of the Basle Capital

Accord in 1988 also spurred the growth of securitization activities. By sharing the risk of

loans, or by taking them off their books, the banks could use the regulatory capital that would

normally have been tied down to cover these loans for other purposes.108 Thus, credit risk

transfer allowed them to diversify their overall portfolios, as well as to transcend their own

capital limitations and tap the broader capital markets to fund their lending activities. As

these techniques extended beyond mortgages to other categories of assets, they gradually gave

107 See Id. at 814ff; Emily R. Pollack, Assessing the Usage and Effect of Credit Derivatives 21ff (April 2003) (Unpublished Paper, Program on International Financial Systems, Harvard Law School); Schuyler K. Henderson, Credit Derivatives, in CREDIT DERIVATIVES: LAW, REGULATION AND ACCOUNTING ISSUES 1, 35-36 (Alastair Hudson ed., 2000) A recent study by the United Kingdom’s Financial Services Authority pointed to the residual risk that some credit derivatives may be recharacterized as insurance: see FINANCIAL SERVICES AUTHORITY, CROSS-SECTOR RISK TRANSFERS, Annex B (2002). 108 See Ian Bell & Petrina Dawson, Synthetic Securitization: Use of Derivative Technology for Credit Transfer, 12 DUKE J. COMP. & INT’L L. 541, 3 (2002) (“The sovereign debt crisis and the Basle accords marked the beginning of a transformation of world banking. From then on, the management of the capital base became one of the key management tasks of major banks.”)

- 39 -

rise to the idea that credit exposures, which had traditionally been held until maturity, could

be seen as tradable commodities.109

Credit derivatives, when they grew in importance in the 1990s, added powerful new

tools to the banks’ arsenal of sophisticated risk management practices. Initially seen as a

complement to loan securitization techniques, credit derivatives rapidly started developing

independently.110 A major advantage of credit default swaps is that they can be purchased to

hedge the risk exposure to a loan without selling the loan itself and potentially jeopardizing

the bank’s ongoing business relationship with the debtor. Thus, a bank with a profitable

relationship with a large debtor can continue to benefit from its business, without creating a

concentrated portfolio that would make the bank excessively vulnerable to changes in that

particular debtors’ creditworthiness.111 In addition, the bank can continue to service the loans

without the additional legal and administrative costs involved by serving as agent under a

securitization structure.112 Finally, while partial loan sales are complex and cumbersome,

credit default swaps can easily be structured so as to transfer only part of the credit risk on a

given loan.113

The benefits of such credit transfers can be further leveraged through more

sophisticated structures, such as synthetic CDOs – a securitization technique under which the

credit risk associated with the securitized asset is transferred to the securitization vehicle

through a credit default swap, rather than by transferring the asset itself. Various tranches of

109 CREDIT RISK TRANSFER, supra note 5 at 4. 110 See Romain Ranciere, Credit Derivatives in Emerging Markets, IMF Policy Discussion Paper (September 2001) at 4. 111 See Bell & Dawson, supra note 108, at 549. 112 Id., at 550. 113 Id.

- 40 -

debt in the securitization vehicle, corresponding to various degrees of risk, are then sold to

third-party investors.

Another widespread technique is the sale of credit-linked notes (CLNs). Under this

structure, the protection seller under a cash-settled CDS turns around and sells CLNs to

investors. Both the original CDS and the CLNs are referenced to the same entity. The CLN

buyers pay an up-front principal amount and earn interest from the seller, just as they would

from regular notes. The payments under the CLNs, however, will be reduced if a credit event

occurs with respect to the reference entity. The protection seller under the CDS has thus

passed the credit risk along to the CLN investors. A significant advantage of CLNs is that,

since they are funded upfront, they eliminate the counterparty risk inherent in regular CDSs –

i.e., the risk that the protection seller may be unable to settle the swap when a credit event

occurs.114 Moreover, since CLNs can be to many investors in relatively small denominations,

their existence improves the liquidity of the credit risk transfer market. Finally, CLNs can be

structured so as to replicate the characteristics of the underlying obligations held by the

original protection buyer, thus providing the market with investment opportunities that would

otherwise be unavailable.115 It will be noticed that both synthetic CDOs and CLNs involve

use of an underlying cash-settled credit default swap, and are normally based on the same

credit event definitions.116 Therefore, the rest of this paper will focus on the characteristics

and documentation of credit default swaps.

114 See Gunter Dufey & Florian Rehm, An Introduction to Credit Derivatives, University of Michigan Business School, Working Paper 00-013 (August 1, 2000), at 6. 115 See Bell & Dawson, supra note 108, at 554. 116 See Jeffrey Tolk, Moody’s Investor Services, Understanding the Risks in Credit Default Swaps (March 16, 2001) at 3-4.

- 41 -

On the demand side, investors such as insurance companies have a substantial interest

in diversifying their own portfolios by acquiring exposure to bank loans and other assets.117

This demand was compounded in the 1990s by a long period of low interest rates. As a result

of these powerful demand and supply forces, a substantial market in credit risk transfer has

developed. Markets participants now include, in addition to banks and insurance companies,

securities dealers (acting either as intermediaries or for their own account), investments funds

(such as pension funds, mutual funds and hedge funds), and government agencies.118 Hedge

funds, in particular, now utilize increasingly elaborate investment techniques based on credit

derivatives to acquire and customize various synthetic exposures.

Finally, CDSs also provide valuable information to the market, by creating a

mechanism through which the credit risk of an asset is priced separately from its other

features.119 For this reason, increases in the spread for CDSs referenced to a particular entity

may foretell future financial difficulties or default.120 This market-based source of information

can be of use, not only to investors, but also to domestic and international policymakers.

2. Sovereign Credit Derivatives

Sovereign credit derivatives, consisting primarily of credit default swaps referenced on

sovereign obligations,121 constitute an important segment of the market. A recent study

117 See generally INTERNATIONAL ASSOCIATION OF INSURANCE SUPERVISORS, IAIS PAPER ON CREDIT RISK TRANSFER BETWEEN INSURANCE, BANKING AND OTHER FINANCIAL SECTORS PRESENTED TO THE FINANCIAL STABILITY FORUM (2003) (hereinafter IAIS PAPER). The IAIS report estimates that insurance companies globally may be accepting around $667 billion in credit derivatives (Id. at 3). 118 CREDIT RISK TRANSFER, supra note 5, at 6-9; Pollack, supra note 107, at 4. 119 See Feder, supra note 98 at 707. 120 See, e.g., Jorge A. Chan-Lau, Anticipating Credit Events Using Credit Default Swaps, with an Application to Sovereign Debt Crises (IMF Working Paper, May 2003); Ranciere, supra note 110 at 9-14; but see Fitch Ratings, Credit Derivatives: A Case of Mixed Signals? (December 4, 2003) on the limitations of this approach. 121 See Ranciere, id., at 5-6 (“[T]he default swap is the cornerstone product in the credit derivatives emerging markets, accounting for 85% of outstanding notional according to Deutsche Bank”). Other instruments, such as

- 42 -

conducted by the Bank for International Settlements on the database maintained by

CreditTrade, a major credit derivatives broker, found that sovereign CDSs accounted for 7.4%

of all quotes in 2002 and 2003.122 While sovereign CDSs were not the fastest-growing

segment, their general movement in volume reflected those of the market, with rapid growth

between 1997 and 1999, a marked slowdown in 2000 and a resumption of growth between

2001 and 2003.123 Unlike CDSs traded on corporates and banks, however, sovereign CDSs

suffered a temporary setback in 2002, as trading in one of the major names, Argentina, ground

to a halt following its late 2001 default.124

The mean number of quotes per name was higher for sovereigns than for other

categories, given the relatively small number of actively traded names. The market is highly

concentrated in a small number of debtors. Brazil, Mexico, Japan, the Philippines and South

Africa accounted for more than 40% of all sovereign quotes between 2000 and 2003.125 More

generally, the market is overwhelmingly concentrated on emerging market debtors, with

90.5% of all quotes.126 This, of course, is unsurprising, given that these debtors generally

present significantly higher credit risk than developed countries. It does, however, raise the

question as to how credit protection can interact – and possibly interfere – with sovereign debt

restructurings in emerging markets.

credit-linked notes, basket CDSs, synthetic CDOs, and CDSs on large emerging market corporate debtors (mostly state-owned enterprises, particularly in the oil sector), also exist. See id. at 6-7; Jane Herring, Credit Derivatives in Emerging Markets: A Product Analysis, in CREDIT DERIVATIVES: APPLICATIONS FOR RISK MANAGEMENT 11 (1998). 122 See Packer & Suthiphongchai, supra note 5, 79 at 81. 123 Id. 124 Id. at 82. 125 Id. 126 Id. at 83. See also Ranciere, supra note 110, at 5, who cites August 2001 Deutsche Bank estimates of the total size of the emerging markets credit derivatives market at $200-300 million, with 50-60% on Latin America, 23-30% on Eastern Europe, and 10-20% on other regions.

- 43 -

The principal market participants are hedge funds, who use sovereign CDSs to gain

leveraged returns through unfunded synthetic exposures, as well as through various arbitrage

techniques. Mutual funds and pension funds are also protection sellers, gaining exposure to

sovereign credits through credit-linked notes and CDOs. They also occasionally buy

protection on their long bond positions. Banks, as indicated above, generally buy protection

to hedge their existing loan or bond exposures. Finally, broker-dealers act as intermediaries,

notably by repackaging synthetic credit exposures as credit-linked notes sold to investors, as

well as between protection buyers and sellers under regular CDSs. They also retain some risk

as part of their own trading books.127

3. Potential Market Failures

Despite the benefits that make them attractive to market participants, credit risk

transfer techniques – including credit default swaps – are not without drawbacks. In a recent

study, the Committee on the Global Financial System reviewed several market failures that

may be associated with credit risk transfer.128 Although these risks were principally discussed

in connection with securitization techniques such as CDOs, they also apply to credit defaults

swaps. Unsurprisingly, they also are very similar to the kinds of market distortions that can

arise under insurance contracts.

First, several potential problems arise out of the asymmetry of information between

the protection buyer and the protection seller. For instance, the bank that has originated a

loan will normally monitor its debtor’s creditworthiness on a continuous basis, and may have

advance warning of impending financial difficulties. Thus, the bank will have an incentive to

127 See Ranciere, id., at 7-8. 128 See CREDIT RISK TRANSFER, supra note 5, at 16ff.

- 44 -

select its lowest quality exposures for credit risk transfer – an adverse selection problem. In

turn, this practice could potentially close the credit risk transfer market to high quality

exposures, through the following mechanism. Since protection sellers have cannot tell the

quality of the credit risk being transferred, they will discount all credit risks accordingly – in

other words, they will charge a higher spread for taking them on. Protection buyers, however,

who can tell which of their credit risks are of high quality, will refuse to transfer them under

these conditions, as the protection will be too expensive in relation to the risk. Therefore,

only low quality risks will be offered, and the protection sellers will further discount

accordingly. As a result, only low quality risks will be traded, at a discounted price, i.e., for

high spreads. This is known as a “market-for-lemons” problem, after George Akerlof’s Nobel

Prize-winning 1970 article.129

The credit risk transfer market has developed several techniques to alleviate this

problem. In the context of CDOs, for instance, this is sometimes accomplished by requiring

the originating institution to randomly select the obligations from its existing portfolio. More

frequently, the institution retains some proportion of the securitization vehicle’s equity (which

absorbs the first losses on the transferred portfolio). Both these techniques tend to align the

protection buyer’s selection incentives with the protection seller’s.

In general terms, asymmetric information is less of a problem for single-name

instruments such as regular credit default swaps, because they are limited to widely-traded

reference entities about which plentiful public information is available. This is particularly

true with respect to CDSs on sovereign obligations, since events affecting sovereign debtors

129 See George Akerlof, The Market for Lemons: Quality Uncertainty and the Market Mechanism, 84 Q.J. ECON. 488 (1970).

- 45 -

are reported and analyzed by multiple media and official sources, so that existing creditors are

unlikely to possess material nonpublic information about the debtor. Therefore, adverse

selection is not likely to be a significant problem in the sovereign CDS market.

Second, the credit risk transfer may affect the behavior of the protection buyer in its

ongoing relationship with the borrower. For instance, the protection buyer will have reduced

incentives to continue monitoring the borrower’s creditworthiness or to take action against

breaches by the borrower of obligations imposed by the original contract.130 This is known as

moral hazard. Once again, this problem can largely be solved by using the techniques

mentioned above.131 In other situations, however, the interests of the protection buyer can be

completely at odds with those of the protection seller. The protection buyer may refuse to

extend further loans to the borrower, whereas it would have done so were it exposed to the

full risk associated with a default.132 Here too, the problem is less acute in the sovereign

market, since most debt is now held in freely tradable bond form and most holders no longer

have ongoing lending or other business relationships with the debtor.

Third, and most importantly, the protection buyer may exploit unanticipated situations

by taking advantage of the credit transfer contract’s failure to provide for all possible

eventualities. This incomplete contracting problem can arise in situations where an event

occurs which, although technically constituting a credit event under a CDS, does not reflect a

decline in the reference entity’s creditworthiness. Issues of this nature have arisen repeatedly

with respect to the “Restructuring” credit event included in many CDSs, as will be discussed 130 See CREDIT RISK TRANSFER, supra note 5 at 18, 21. 131 See id. at 18. 132 However, this effect is not free from ambiguity. See CREDIT RISK TRANSFER, supra note 5, at 22: “[Government-owned companies in emerging markets] had in some cases perceived a change in their relationship with those banks that were active users of CRT, with expanded credit lines apparently being made available to them.”

- 46 -

in Part III. This problem, of course, will be compounded if the protection buyer actually

possesses the capacity to provoke, or increase the likelihood of, a default on the part of the

reference entity. Thus, one of the fundamental questions raised by the use of sovereign credit

derivatives is whether the legal framework that governs them provides adequate legal

certainty.

C. The Documentation of Credit Derivatives

1. ISDA Documentation

Although parties to OTC derivative transactions are free to document these

instruments as they wish, the International Swaps and Derivatives Association (ISDA) – an

industry group based in New York – has developed a widely-used set of standard

documentation. ISDA-led standardization has been a major factor in the rapid development

of OTC derivatives markets,133 and ISDA has become an important de facto regulator of OTC

derivatives.134

Parties to OTC derivatives typically enter into an ISDA Master Agreement, which

provides a framework for their legal relationship. This standardized contract includes many

boilerplate provisions, such as the representations and covenants of each party, the netting and

calculation of payments, procedures applicable upon default or early termination, as well as

provisions concerning notices, transfer, contractual currency, governing law and jurisdiction,

and tax arrangements.135 The parties can also customize the Master Agreement by choosing

133 See id. at 4. 134 See Frank Partnoy, ISDA, NASD, CFMA and SDNY: The Four Horsemen of Derivatives Regulation (USD School of Law, Public Law and Legal Theory Working Paper 39). 135 See ISDA, 1992 ISDA MASTER AGREEMENT (1992) and ISDA, USER’S GUIDE TO THE 1992 ISDA MASTER AGREEMENT (1992). ISDA published a revised Master Agreement in 2002, which retains the same basic structure. The 1992 Master Agreement remains widely used in the industry.

- 47 -

various options in an attached Schedule, or by amending the Agreement itself. Within this

framework, the terms of individual trades are documented by Confirmations. In case of

inconsistency, the Confirmation overrides the Master Agreement and the Schedule with

respect to that trade. This two-tiered structure has many advantages, including allowing

parties to net the payments due under various trades, and providing an orderly mechanism for

unwinding transactions upon the default or bankruptcy of one of the parties.136

Confirmations describe the essential economic terms of the individual trade through

the use of definitions published by ISDA with respect to major types of OTC derivatives

transactions. Thus, credit default swaps are typically documented by reference to the 1999 or

2003 ISDA Credit Derivatives Definitions.137 The 2003 Definitions preserve the structure

adopted by ISDA in 1999, with some modifications. Both sets of definitions allow the parties

to set out the basic terms of their trades by choosing among a menu of options. Once again, in

case of inconsistency, the Confirmation overrides the standardized Definitions. As a result,

the parties are free to customize the standard ISDA documentation as they see fit, although

such an exercise can be somewhat perilous for inexperienced practitioners.

The principal terms covered are as follows. The parties must choose an Effective Date

and a Scheduled Termination Date, which will determine the life of the swap. They also

designate the Reference Entity – the entity with respect to which a Credit Event will trigger

payment under the swap. Crucially, they also designate the acceptable Categories and

Characteristics of two sets of obligations. The first set – defined simply as Obligations – are

136 Section 2(c) of the 1992 Master provides for netting of amounts due under the same transaction, and allows the parties to designate groups of transactions for netting. 137 See 1999 ISDA CREDIT DERIVATIVES DEFINITIONS (1999) [hereinafter the 1999 DEFINITIONS]; 2003 ISDA CREDIT DERIVATIVES DEFINITIONS (2003) [hereinafter the 2003 DEFINITIONS].

- 48 -

the ones with respect to which a credit event will trigger the swap. For instance, suppose A

purchases a CDS from B, with Brazil as a Reference Entity, Bond or Loan as the applicable

Obligation Category, and Not Domestic Currency as one of the applicable Obligation

Characteristics. Under this swap, B will have to pay A if a credit event occurs with respect to

bonds or loans issued by Brazil. However, B will not have to pay if the credit event relates to

bonds or loans denominated in Brazilian reals. Thus, by carefully choosing the applicable

Category and Characteristics, the parties can tailor their swap to the kinds of risk they deem

relevant.138

The second set – defined as Deliverable Obligations – are the obligations that the

protection buyer will be entitled to deliver for payment under a physically-settled CDS. Thus,

this choice does not apply to a cash-settled CDS. Deliverable Obligations are also determined

by the choice of a Category and of one or many Characteristics. Suppose, under the example

above, that the swap is a physically-settled one, and that the parties have chosen the same

Category and Characteristics for Deliverable Obligations as they did for Obligations. A

default by Brazil on its foreign currency-denominated bonds will trigger the swap, since the

bonds are within the applicable Category and are not denominated in reals. If A, however,

happens to own real-denominated bonds, she will not be allowed to deliver them to B for

payment under the swap, since they do not fit the chosen Deliverable Obligations

Characteristics. This is true whether or not the real-denominated bonds have also been

defaulted on by Brazil.

138 Under the 1999 and 2003 Definitions, the parties must choose one Obligation Category among Payment, Borrowed Money, Reference Obligations Only, Bond, Loan or Bond or Loan. They may also choose one or more of the following Characteristics: Pari Passu Ranking (which became “Not Subordinated” in 2003), Specified Currency, Not Sovereign Lender, Not Domestic Currency, Not Domestic Law, Listed and Not Domestic Issuance.

- 49 -

Emerging market CDSs typically have a broader set of Deliverable Obligations than

Obligations, so as to avoid failures to deliver.139 As a result, the protection buyer often

benefits from a “cheapest-to-deliver” option upon default: she can maximize the economic

benefit of the swap by delivering the cheapest available Deliverable Obligations.140 Thus, the

restrictions on Deliverable Obligations have a crucial impact on the potential payoff from the

swap.

In addition to defining the relevant obligations, the parties must also select the

applicable Credit Events. Under the Definitions, the choices are: Bankruptcy,141 Obligation

Default, 142 Obligation Acceleration, 143 Failure to Pay, 144 Repudiation/Moratorium 145 and

139 See Ranciere, supra note 110, at 15. 140 Id. 141 Bankruptcy occurs when the Reference Entity: (a) is dissolved; (b) becomes insolvent or admits it in writing; (c) makes a general assignment, arrangement or composition for the benefit of its creditors; (d) institutes or has instituted against it a proceeding seeking a judgment of insolvency, bankruptcy or another similar law affecting the rights of creditors, and such proceeding (i) results in such a judgment; or (ii) is not dismissed within 30 days; (e) has a resolution passed for its winding-up, official management or liquidation; (f) seeks or become subject to the appointment of a receiver or a similar officer; (g) has substantially all of its assets taken possession of by a secured creditor; (h) causes or is subject to any similar event; or (i) takes any action in furtherance, or consents to, any of the foregoing acts. This overview omits many of the details in the actual definition: see 1999 DEFINITIONS, s. 4.2; 2003 DEFINITIONS, s. 4.3. 142 Obligation Default occurs when “one or more Obligations have become capable of being declared due and payable…as a result of…the occurrence of a default [or] event of default…other than a failure to make any required payment.” [emphasis added]. 1999 DEFINITIONS, s.4.4; 2003 DEFINITIONS, s. 4.4. This credit event encompasses many technical defaults that do not reflect upon the Reference Entity’s creditworthiness and is rarely used. 143 Obligation Acceleration covers the same type of event as Obligation Default, with the difference that, in addition to the Obligations being capable of being declared due and payable, they have actually been so declared. 1999 DEFINITIONS, s. 4.3; 2003 DEFINITIONS, s. 4.3. 144 Failure to Pay means “after the expiration of any applicable (or deemed) Grace Period…, the failure by a Reference Entity to make, when and where due, any payments in an aggregate amount of not less than the Payment Requirement under one or more Obligations.” 1999 DEFINITIONS, s. 4.5. The 2003 DEFINITIONS add: “in accordance with the terms of such Obligations at the time of such failure.” 2003 DEFINITIONS, s. 4.5. The Payment Requirement is specified by the parties in their confirmation; the default amount is USD 1,000,000. 1999 DEFINITIONS, s. 4.8(d); 2003 DEFINITIONS, s. 4.8(d). 145 Repudiation/Moratorium occurs when “a Reference Entity or Governmental Authority (a) disaffirms, disclaims, repudiates or rejects, in whole or in part, or challenges the validity of, one or more Obligations in an aggregate amount of not less than the Default Requirement or (b) declares or imposes a moratorium, standstill or deferral, whether de facto or de jure, with respect to one or more Obligations in an aggregate amount of not less than the Default Requirement.” 1999 DEFINITIONS, s. 4.6. The Default Requirement is set by the parties, with a default amount of USD 10,000,000. 1999 DEFINITIONS, s. 4.8(a). It is unclear what clause (b) adds that

- 50 -

Restructuring.146 Upon the occurrence of one of the selected events, the protection buyer must

deliver to the protection seller a notice containing “a description in reasonable detail of the

facts relevant to the determination that a Credit Event has occurred.”147 If the confirmation so

provides, the protection buyer must also deliver a notice citing publicly available information

confirming such facts.148 Once these requirements are met, the protection seller must make

payment according to the chosen settlement method. Finally, the Confirmation also provides

for the amount of the premium paid by the protection buyer, dispute resolution mechanisms,

and various technical details.

2. Credit Events and the Restructuring Debate

The main problem affecting the documentation of sovereign credit derivatives has

been defining the applicable credit events. This problem also affects other credit derivatives.

It does so, however, to a lesser degree, because the failure of corporate entities is subject to

well-defined legal regimes such as bankruptcy and restructuring laws. 149 Even in the

corporate context, however, difficulties remain.150 In recent years, the principal example of

wouldn’t be covered by the Failure to Pay credit event. Clause (a), however, might cover anticipatory repudiations, although no Failure to Pay has yet occurred. The Repudiation/Moratorium definition has undergone considerable clarification in the 2003 DEFINITIONS (s. 4.6). The new definition only applies when the repudiation or moratorium is actually followed within 60 days (or on the next payment date for bonds) by a Failure to Pay or Restructuring (regardless of the amount). 2003 DEFINITIONS, s. 4.6. 146 The Restructuring clauses of both the 1999 DEFINITIONS and the 2003 DEFINITIONS are reproduced in Appendix 1 and discussed in detail, infra. 147 1999 DEFINITIONS, s. 3.3. The parties may also allow the protection seller to originate the Credit Event Notice by designating “Buyer and Seller” as the Notifying Parties. See s. 3.2(b)(ii). 148 This is called a Notice of Publicly Available Information. “Publicly Available Information” is in turn defined as information published in a previously agreed number of internationally recognized news sources, information received from the Reference Entity or its agent or trustee, and information filed in certain legal proceedings or stated in a judgment or other court or regulatory order. See 1999 DEFINITIONS, s. 3.5, for details. 149 In the United States, both regimes are embodied in the United States Bankruptcy Code, 11 U.S.C. 101ff. In some foreign jurisdictions, corporate restructurings are dealt with under a legal regime distinct from bankruptcy. 150 CREDIT RISK TRANSFER, supra note 5 at 19.

- 51 -

the uncertainties associated with the ISDA Credit Events has been the debate surrounding the

Restructuring definition.

As mentioned earlier, credit risk is the risk that a debtor will default on an obligation

due to a deterioration in its creditworthiness. Market participants who trade in credit

derivatives expect these instruments to transfer credit risk separately from market risk. There

arise, however, many cases where distinguishing one from the other is a challenge.

Nevertheless, market participants attempt to do so by isolating events that genuinely reflect a

deterioration in the reference entity’s creditworthiness from mere market movements.

Despite the important distinctions between credit derivatives and insurance, an important

indicator of credit risk is whether holders of the obligation actually suffer a loss as a result of

the event. For this reason, Obligation Default is almost never included by parties as a Credit

Event, because many technical defaults on debt obligations – for instance, a failure to deliver

financial reports on time, or a minor deviation from a restrictive covenant – do not result in

any loss on the part of the creditors, and there occurrence is not a materialization of credit

risk.151 Of course, separating credit risk from market risk cannot be achieved with perfect

accuracy, but the exercise is useful insofar as it makes credit protection activities more

reliable and less costly.

From this perspective, determining which restructurings should constitute credit

events raises thorny issues. For instance, many debt renegotiations resulting in change to the

terms of the underlying obligation do not mean that the debtor is encountering financial

difficulties. On the contrary, an agreement on more favorable lending terms may be the result

of an improvement in the debtor’s credit, and thus in its negotiating position. Moreover, the 151 See Bendernagel, supra note 106, at 280; see also Tolk, supra note 116, at 9.

- 52 -

variety of possible debt renegotiation scenarios – from informal negotiations with individual

creditors to full-fledged restructuring under Chapter 11 or analogous foreign laws – makes

precise definition of the underlying event itself problematic.152

Since the 1999 Definitions were adopted, ISDA’s approach to Restructuring has been

as follows. First, the definition requires that the change in the obligation fall within a list of

adverse changes: a reduction in interest or principal, a postponement of their accrual or

payment, a change in ranking or priority of the obligation, or a change in the currency or

composition of any payment.153 The 1999 Restructuring clause expressly encompasses the

occurrence of such changes as a result of an Obligation Exchange, defined as the mandatory

transfer of any securities, obligations or assets to holders of Obligations in exchange for such

Obligations. As will be seen, this requirement played a crucial role in the litigation

surrounding Argentina’s 2001 pre-default exchange, and was dropped from the 2003

Definitions. Second, the event must affect the Obligations in an aggregate amount no less than

a Default Requirement amount predetermined by the parties.154 Third, the event must not be

provided for under the terms of the Obligation.155 Finally, the definition excludes certain

technical adjustments, as well as events that do not “directly or indirectly result from a

deterioration in the creditworthiness or financial condition of the Reference Entity.”156

Limiting the Restructuring definition and other credit events in this way, of course, is

primarily of interest to protection sellers. Definitions that encompass events that do not truly

cast aspersions on the reference entity’s creditworthiness would allow protection buyers to

152 See Tolk, id., at 8. 153 See 1999 DEFINITIONS, s. 4.7(a). 154 See id. If the parties do not specify an amount, the Default Requirement is deemed to be USD 10,000,000. 155 See id. 156 See id., 4.7(b)(iii)

- 53 -

trigger the swaps opportunistically and obtain full payment of the reference amount – which

may be higher than the market value of the obligations at that time. Thus, broad credit events

may amount to free protection against market risk for the protection buyer.157 Conseco’s

2000 restructuring illustrated this problem. The troubled insurer reached an agreement with

its lender banks to restructure its credit facilities. The agreement technically constituted a

Restructuring under the 1999 Definitions, because the maturity for one short-term loan had

been extended. Accordingly, many protection buyers triggered their swaps.158 In addition,

they exercised their “cheapest-to-deliver” option by delivering long-term Conseco securities,

whose value had declined precipitously following the restructuring, to a level well below that

of the bank debt that was actually restructured.159

Although the Conseco swaps appear to have been settled without litigation, many

protection sellers considered the result inappropriate. 160 ISDA subsequently adopted an

optional Modified Restructuring Definition that excluded restructurings affecting Obligations

held by less than three holders, or adopted by consent of less than two-thirds of holders.161

With respect to physically-settled derivatives, the modified definition also prohibited

protection buyers from delivering obligations with a maturity date more than thirty months 157 This is a crucial concern for credit rating agencies hired to rate CLNs or synthetic CDOs: see Tolk, supra note 116 at 5. For instance, if the credit event definitions are broader than the rating agency’s definition of a default, then the risk associated with the CLN or CDO will in fact be greater than indicated by the rating calculated on the basis of the probability of a “real” default by the reference entity. The rating agencies’ desire to capture only substantive (as opposed to formal or technical) defaults, however, can clash with the certainty and rapidity of execution sought by protection buyers (Id. at 8). For instance, Moody’s definition of default, instead of focusing on specific aspects of a restructuring, involves an evaluation of whether the process results in a “diminished financial obligation,” or “has the apparent purpose of helping the borrower avoid default” (Id. at 5) The substantial subjective aspects of such an evaluation have prevented its adoption in ISDA forms. 158 See Pollack, supra note 107, at 28. 159 See Ranciere, supra note 110, at 18. 160 See HENDERSON ON DERIVATIVES, supra note 98, at 149; Pollack, supra note 107, at 28-30. 161 See ISDA, RESTRUCTURING SUPPLEMENT TO THE 1999 CREDIT DERIVATIVES DEFINITIONS, s. 4.10 (2001); 2003 DEFINITIONS, s. 4.9; The idea here is that widely-held obligations are less likely to be restructured for non-credit reasons: see Tolk, supra note 116, at 8. The Supplement does not seem to have been widely adopted in the sovereign CDS market; see Packer and Suthiphongchai, supra note 5, at 83.

- 54 -

after that of the restructured obligation or the termination date of the swap, whichever is

earlier.162 Despite these amendments, the Conseco experience, along with a similar one

following Xerox’s restructuring of its bank loans,163 led several protection sellers to argue for

the outright elimination of the Restructuring definition from ISDA documentation.

In addition to accurately separating credit risk from market risk, another goal of the

definition-setting exercise, from the protection sellers’ perspective, is to reduce moral hazard

by avoiding situations where the buyer will have the incentive and capacity to precipitate a

credit event. This objective generally coincides with a broader systemic interest in

encouraging debtors to participate in restructurings. The modified Restructuring definition

took a step in that direction by excluding modifications to smaller classes of Obligations,

which are more vulnerable to manipulation by protection buyers.

The interests of protection sellers in avoiding moral hazard, however, are not the only

ones at play. Protection buyers who use credit derivatives to hedge their exposures to credit

risk need these instruments to deliver adequate protection. Courts have recognized this

legitimate interest. In Ursa Minor Limited v. Aon Financial Products,164 the Second Circuit

held that the broad waiver of defenses in a credit default swap prevented protection sellers

from asserting defenses based on the alleged invalidity of the government-backed security

bond whose default had triggered the swap. In another case, however, a technical

interpretation prevailed and resulted in a windfall to protection buyers. In Deutsche Bank AG

v. ANZ Banking Group Ltd.,165 the Commercial Division of the Queen’s Bench held that,

162 See RESTRUCTURING SUPPLEMENT, id., s. 2.29; 2003 DEFINITIONS, id., s. 2.32-2.33. 163 See CREDIT RISK TRANSFER, supra note 5 at 20. 164 7 Fed. Appx. 129 (2d Cir. 2001), aff’g 2000 WL 1010278 (S.D.N.Y.); see Pollack, supra note 107, at 7. 165 2000 WL 1151384 (QBC (Comm. Ct.))

- 55 -

since the obligations of the City of Moscow did not provide for a grace period, a default that

lasted only one day had triggered a credit event.166 As a result, Deutsche Bank, the protection

seller, had to accept the obligations and settle the swap, effectively providing ANZ with

market risk protection it had not bargained for.

Despite such occasional excesses, the need of protection buyers for adequate

protection is particularly acute when banking regulators allow banks to use credit derivatives

as risk mitigation techniques and correspondingly to reduce their capital requirements.167

Credit event definitions so narrow that they exclude events that cause genuine loss to

protection buyers might destabilize the banking system. This is why, for instance, many

banking regulators insist that credit derivatives used by banks to hedge credit risk include

Restructuring as a Credit Event. European banking authorities have been adamant on this

point,168 and the current Basel proposals do not contemplate capital relief for hedges that

exclude Restructuring as a credit event.169

Notably, despite its increasing disrepute among market participants, the Restructuring

credit event has remained an essential feature of virtually all sovereign credit derivatives.170 It

is easy to see why. Under the current sovereign debt regime (or absence thereof), countries

166 In the aftermath of this dispute, ISDA adopted a three business days minimum grace period requirement, which applies even when the underlying obligation does not provide for a grace period. See 1999 DEFINITIONS, s. 1.11. “The purpose of the concept is to reduce the likelihood that, with respect to Obligations having no grace period or a very short grace period, a Failure to Pay Credit Event will be triggered by a failure to make a relevant payment for administrative reasons or other reasons that do not reflect upon the creditworthiness or financial condition of the Reference Entity.” (Id. at vii). This provision has been maintained in the 2003 DEFINITIONS, s. 1.12. 167 See CREDIT RISK TRANSFER, supra note 5, at 20; HENDERSON ON DERIVATIVES, supra note 98, at 150. 168 See Pollack, supra note 107, at 31. 169 See CREDIT RISK TRANSFER, supra note 5 at 20; Pollack, id., at 37-42; INTERNATIONAL FINANCE, supra note 10, Draft ch. 15, at 20-21. 170 Ranciere, supra note 110, at 16: “Restructuring is very often excluded from corporate credit derivatives but is always included as a credit event in sovereign credit derivatives”. See also Packer & Suthiphongchai, supra note 5, at 83.

- 56 -

do not go bankrupt. They also very rarely fail to pay: outright sovereign debt defaults like

Russia’s or Argentina’s are the exception, not the rule. Nevertheless, sovereigns frequently

encounter severe economic difficulties that make it impossible to continue servicing their

current debt level. The normal outcome of such difficulties is the negotiation of a “voluntary”

debt restructuring with the countries’ external creditors.171 Such restructurings often reflect a

substantial deterioration of the sovereign’s creditworthiness and, as such, are a manifestation

of credit risk. Thus, absent the Restructuring definitions, the protection offered by sovereign

credit derivatives would likely be inadequate.

In spite of the importance of the Restructuring definition to the sovereign swap

market, this clause, like the other standard credit events, was developed in the corporate

context and does not consistently reflect sovereign market concerns.172 In addition, while

parties may adapt the definitions, and doubtlessly do so in highly customized transactions, the

available evidence shows that most sovereign CDSs still incorporate the unmodified 1999

Restructuring definition.173 Thus, more recent amendments to that provision may not be

particularly relevant to the current sovereign credit default swap market. This raises questions

as to whether the current credit event definitions can handle sovereign debt restructurings

adequately, and whether the results coincide with broader systemic interests in preventing

holdout litigation and, ultimately, promoting successful restructurings as a debt reduction

technique for indebted countries.

171 See Martin Hughes, Areas of Legal Risk in Sovereign-Linked Credit Derivatives, in CREDIT DERIVATIVES: LAW, REGULATION AND ACCOUNTING ISSUES, supra note 107, 64 at 65-66. 172 See Id., at 65. 173 See Packer & Suthiphongchai, supra note 5, at 83 (“Unlike CDSs written on bank and corporate obligations, the vast majority of outstanding sovereign CDSs remain governed by the old restructuring clause of the 1999 ISDA Credit Derivatives Definitions.”) The proportion appears to be in excess of 95%.

- 57 -

III. CREDIT DERIVATIVES AND THE RESTRUCTURING PROCESS

Under the current restructuring process, distressed sovereigns and their creditors

conduct negotiations on a case-by-case, voluntary basis. Restructurings typically fall within

one of two categories, depending on whether they occur before or after a formal default on the

affected debt.174

A. Post-Default Restructurings

Post-default restructurings were the norm in sovereign finance during the 1980s and

early 1990s. The sovereign would first suspend payments on its debt, usually in the midst of a

serious economic crisis. There ensued a period during which the creditors received reduced

payments of principal or interest, or no payments at all. As a result of the suspension, a

majority of creditors could accelerate the debt, but they had very limited enforcement options

beyond this. Instead, they often chose to avoid formally declaring a default. The sovereign,

for its part, was barred from further access to the international financial markets, which

hindered economic growth and government activities.

Eventually, the parties would agree to initiate negotiations towards a debt

restructuring. From the sovereign’s perspective, the principal benefit of a composition was

renewed access to international capital. Provided that future growth prospects were such that

the return from new capital inflows would exceed the cost of resuming payments on a reduced

schedule, the sovereign had an incentive to split that surplus with the creditors in return for

bringing the default to an end. The creditors, for their part, would accept the restructuring

174 See Bratton & Gulati, supra note 78, at 18-19 (observing that, although “liquidity crises often move more quickly than the adjustment processes and default proves unavoidable,” several countries, including Pakistan, Ecuador, Ukraine, Argentina and Turkey negotiated pre-default restructurings in recent years).

- 58 -

terms to the extent that they were better than what they could obtain through any available

enforcement techniques. Beyond this, the legal provisions governing the debt – particularly

unanimous voting procedures – along with various extraneous factors (such as official sector

involvement) would determine the strength of the creditors’ negotiating position, and thus,

how much of the surplus they could extract from the debtor.

Once an agreement was reached, the renegotiated terms would either be incorporated

in the existing obligations through a creditor vote, or embodied in new obligations distributed

through an exchange offer. This is how Brady restructurings were typically structured.

Following the exchange, holdout creditors – if any – held a majority of the remaining debt.

They could thus accelerate future payments and sue the debtor. Their enforcement options,

however, were still deficient. Thus, the quasi-totality of creditors typically participated in

these restructurings.

Post-default restructurings do not pose particular difficulties with respect to sovereign

credit derivatives. The initial suspension or default clearly constitutes a Repudiation/

Moratorium or Failure to Pay credit event, so that any swaps referenced to the affected debt

will be triggered at the outset. The sovereign’s deliverable obligations will thus be transferred

to the protection sellers. As a result, the restructuring negotiations will not suffer from the

dislocation of incentives and moral hazard problems associated with credit protection.

B. Pre-Default Restructurings

1. The Exchange Offer Framework

While post-default restructurings of this sort still occur – as the current negotiations

involving Argentine bonds demonstrate – they no longer dominate sovereign finance. In

- 59 -

recent years, a series of successful pre-default restructurings have considerably changed the

landscape. Countries such as Ecuador, Uruguay and Pakistan have restructured their debts

through voluntary exchange offers without having to resort to formal suspension or default.

Distressed sovereign debtors now frequently engage in negotiations with their creditors when

their obligations become unsustainable, but well before a true crisis occurs.

There are several reasons for this. First, when a sovereign’s debt becomes

unsustainable, both the sovereign and its creditors have an interest in avoiding the

consequences associated with a default followed by a protracted suspension of payments.

Second, the official sector undoubtedly prefers an orderly exchange offer to the disruption

caused by a default. Although large official aid packages have accompanied recent pre-

default restructurings, this pattern reduces the need for these packages to be negotiated in

crisis circumstances. It also arguably reduces their size, in comparison with the amounts that

would be needed to reverse a widespread loss of market confidence and a potential currency

crisis. Third, the development of exit consents has provided debtors with a powerful tool to

coax recalcitrant creditors into accepting an exchange offer.

The contemporary pre-default restructuring process, although it is also conducted on a

case-by-case basis, is based on the following pattern. During negotiations, the creditors

cannot accelerate or sue the debtor, since it is still making payments. All parties, however,

know that the sovereign is financially distressed. From the sovereign’s perspective, a default

would have one principal cost: closing its access to international capital. However, it would

also have a benefit: allowing the sovereign to halt payments on its existing debt (except to the

limited extent that creditor enforcement techniques can create additional costs). The benefit

must be higher than the cost, otherwise the sovereign would not be contemplating default.

- 60 -

Thus, the sovereign will rationally want to make an offer that is between what the creditors

expect to receive after default, and the full value of continued access to international markets.

Where exactly the line will be drawn depends, once again, on the respective negotiating

strength of the parties and on extraneous factors, such as official sector intervention.

In practice, the restructuring is implemented through an exchange offer. The

sovereign makes a public offer to exchange the old bonds for new bonds with less favorable

terms, such as a lower interest rate or a longer maturity. In many cases, the debtor also uses

two additional techniques to provide additional incentives for the bondholders to participate.

First, holders who subscribe to the offer are asked to provide exit consents to allow

amendments that will make the old bonds unattractive. Second, the debtor represents (or at

least strongly intimates) that it will cease payments on untendered old bonds and will not offer

any better terms to non-participating holders. Both these features of exchange offers, of

course, make them quite coercive and weaken the creditors’ position. In some recent cases,

there have been little or no negotiations – the sovereign simply made a ‘take it or leave it’

offer.

Market experience with recent exchange offers indicate that the threshold participation

rate for success is about 90%. In recent restructurings in Ukraine and Ecuador, the

participation rate was over 95%.175 Following a successful exchange, the debtor has to decide

whether to continue servicing any untendered bonds. If it chooses to default, holdout

175 See Gray, supra note 95, at 225. While there is no formal participation threshold, the debtor normally insists on a high participation rate, principally because (i) of the possibility of legal action by holdouts, whatever the chances of success; and (ii) it wants to retain the option to keep servicing the untendered debt while successfully reducing its debt burden. Of course, while the debtor wants to keep the latter option, it will nevertheless attempt to convince creditors that servicing the old debt will not be a priority, so as to encourage them to participate.

- 61 -

creditors are entitled to accelerate and sue. Once again, however, their enforcement clout is

limited by the lack of attachable assets and the costs associated with litigation.

2. Exchange Offers under the 1999 ISDA Definitions

How is a voluntary exchange treated under the 1999 ISDA Definitions? The new

financial terms implemented by an exchange clearly fall within the listed events in Section 4.7

(reduction in the rate or amount of interest or principal, or postponement of payments of

principal or interest). Assuming that the changes affect securities that qualify as Obligations,

the question thus becomes whether such event occurs in the way described in the definition.

First, the paragraph expressly includes events that occur “as a result of an Obligation

Exchange.” Section 4.9 goes on to define an Obligation Exchange as a “mandatory transfer

… of any securities, obligations or assets to holders of Obligations in exchange for such

Obligations.” Second, as an alternative to an Obligation Exchange, the paragraph also

includes events which simply “occur” or are “agreed between the Reference Entity … and the

holders.”176

In the only published case on point, Eternity Global Master Fund v. Morgan Guaranty

Trust Co., the Southern District of New York held that a voluntary debt exchange did not

trigger the Restructuring credit event. Eternity, which had invested in Argentine debt, entered

into three credit default swaps for hedging purposes. The protection seller was Morgan, the

Reference Entity was the Republic of Argentina, and the swaps included both

Repudiation/Moratorium and Restructuring among the applicable Credit Events.

176 See Appendix 1 for the full definition.

- 62 -

On November 1, 2001, the President of Argentina issued a decree instructing the

Minister of the Economy to offer a voluntary exchange to holders of Argentine government

debt. On November 19, Argentina officially announced the exchange. Under its terms, any

tendered debt would be held in trust for the holders, who would be issued new secured debt

obligations with lower yields and longer maturities. This effort, however, proved insufficient

to resolve the crisis. On December 24, 2001, Argentina suspended payments on its debt

obligations.

Starting on November 1st, Eternity repeatedly requested that Morgan settle the swaps.

In response, Morgan took the position that the debt exchange did not constitute a credit event.

Only on December 27, following Argentina’s suspension of payments, did Morgan declare

that a “Repudiation/Moratorium” credit event had occurred. Eternity subsequently sued

Morgan for breach of contract, fraud and misrepresentation. The breach of contract claim

alleged that Argentina’s voluntary debt exchange constituted a Restructuring credit event, so

that Morgan should have settled the swaps in November.

Morgan argued that since the definition of “Obligation Exchange” requires that the

exchange be mandatory, Argentina’s voluntary exchange did not qualify and thus could not

result in a Restructuring. Although the court denied Morgan’s first motion to dismiss,177 it

eventually accepted this argument on a subsequent motion to dismiss Eternity’s amended

complaint.178 First, the court held, since the exchange was not mandatory, it could not be an

Obligation Exchange. Second, even though Eternity had in fact tendered its debt in exchange

for lower-interest, longer-maturity Argentine bonds, this did not mean that a change in the

177 2002 WL 31426310 (S.D.N.Y.) 178 2003 WL 21305355 (S.D.N.Y.)

- 63 -

obligations had been agreed between Argentina and the holder. Under the terms of the

exchange, the obligations held in trust for the holders remained unchanged, so there was no

postponement, deferral or delay in payments or reduction in interest on them.

While the second part of McKenna J.’s reasoning might strain the reader’s credulity, it

is easy to see why he accepted Morgan’s argument. Argentina’s exchange was (at least

conspicuously) voluntary. Surely, it would have seemed odd to allow a holder who

voluntarily agreed to accommodate its creditor, to then turn to its insurer and claim payment

under the swap. Since the holder would not have to live with the restructured obligations, it

might be willing to accept a unfavorable exchange offer. Moreover, allowing a voluntary

exchange to trigger the swap would bypass the clear requirement in the Obligation Exchange

definition that the exchange be mandatory.

3. Effect on Incentives

The reasoning in Eternity is based on the existence of the trust, an uncommon

restructuring technique that was used for reasons peculiar to the Argentine exchange.179 The

result reached in Eternity, however, will likely be followed in future cases involving voluntary

restructurings, regardless of whether the debtor uses the trust device. As a matter of

contractual interpretation, allowing a voluntary exchange to trigger the swap would

circumvent the express requirement that an Obligation Exchange be ‘mandatory.’ To avoid

this result, a court faced with an exchange offer without a trust could argue that the changes in

payments or maturity did not occur “with respect to” the old bonds: the old bonds still have

179 The failed 2001 exchange targeted Argentina’s domestic currency debt (whether held by domestic creditors or foreigners), and was meant to be followed in early 2002 by an exchange offer for Argentina’s foreign currency bonds. Argentina adopted the trust structure in order to avoid concentrating the domestic currency debt in the hands of foreign holdouts from the first exchange, at least until the second exchange could be completed. See e-mail from Jeremiah Pam, on file with the author.

- 64 -

the same payment terms. The new bonds, for their part, have different payment terms, but

those have never been changed. All that has happened is that old bonds have been voluntarily

exchanged for new bonds. Thus, neither the definition of Obligation Exchange nor the

general language of Section 4.7(a) apply. As will be described below, such an interpretation

would be buttressed by the market’s general acquiescence to the Eternity ruling, whose

reasoning was not repudiated by ISDA when it adopted the 2003 Definitions.

Assuming that the Eternity ruling is generally applicable to voluntary restructurings,

how does it affect the incentives of the protected creditor faced with an exchange offer?

Consider first the incentives of the unprotected bondholder (i.e., the bondholder who has not

purchased a credit default swap), as represented by Figure 1. If he participates in the

exchange, he will receive new bonds with a value lesser than that of his old bonds. If he

refuses to subscribe to the offer, however, two outcomes are possible. First, if enough

bondholders similarly hold out, the exchange will fail to attract enough subscriptions (about

90%) and the restructuring will collapse. As a result of the restructuring’s failure, the debtor

will likely default at some point in the future. The bondholders, individually or collectively,

may then try to enforce their claims against the debtor, but the prospects for significant

recovery in such circumstances are uncertain. They may also resume negotiations in order to

achieve a post-default restructuring, but the suspension period will involve serious losses both

to the bondholders – as payments on the bonds will have been suspended – and to the debtor –

whose access to the international financial markets will be shut. Second, the exchange might

reach the 90% threshold despite the non-participation of our individual bondholder. In this

case, he will be left with old bonds after most other creditors have switched to new bonds.

- 65 -

After the debtor defaults on the old bonds, the bondholder will have some remedies, but their

effectiveness will likely be limited, and they will come at substantial transaction costs.

FIGURE 1

Under these circumstances, the bondholder will choose to participate if the present

value of the new bonds is greater than the expected payoff from any available enforcement

methods against the debtor. As mentioned above, the sovereign debtor will calibrate its offer

accordingly. The available enforcement methods, along with other factors affecting the

bondholders’ bargaining power, will determine the proportion of the surplus that they can

extract from the debtor. Once the debtor meets the reservation price of 90% or more of the

debtors, however, the exchange succeeds.

- 66 -

The holder who has bought a credit default swap on his bonds, for his part, faces very

different incentives. As shown by Figure 2, if he refuses to tender his bonds in the exchange

offer, two outcomes are possible. First, it may be that, despite his non-participation, the

exchange will attract enough bondholders to ensure its success. In this case, the non-

participating holder retains the old bonds, on which the debtor subsequently defaults. The

default triggers the credit default swap (as a Failure to Pay or Repudiation), and the holder is

entitled to deliver the old bonds for their face amount to the protection seller. The protection

seller, for his part, obtains highly devalued bonds.

FIGURE 2

Second, if enough bondholders refuse to subscribe to the offer, the exchange fails.

Assuming that the offer was motivated by the debtor’s impending inability to service its debt,

- 67 -

the debtor will eventually default. Once again, upon default, the CDS is triggered and the

protected holder can deliver the bonds for their full face amount. The protection seller obtains

highly devalued, recently-defaulted bonds. On the other hand, if the debtor does not default,

then the protected holder again escapes any sacrifice and keeps receiving payments on the

bonds’ original terms.

Thus, in all cases, the protected bondholder who refuses to subscribe to the exchange

offer retains the full value of the old bonds. If he participates in the exchange offer, however,

he receives new bonds with terms inferior to those of the old bonds.180 Since a voluntary

exchange does not trigger the CDS, he receives no compensating payment from the protection

seller and thus loses the benefit of the swap.181 Moreover, participating in restructuring

negotiations will involve certain costs, while standing aside and waiting for the exchange to

fail costs nothing. Under these conditions, the protected holder will rationally choose to

refuse the offer and pass the costs imposed by the restructuring (or its failure) along to the

protection seller.

Thus, all protected holders should rationally refuse to participate. Their non-

participation increases the likelihood that the exchange offer will not reach the 90% threshold

180 Unless, of course, the exchange fails despite his participation. In that case, he is entitled to full payment under the swap. However, if he holds out, payment under the swap is virtually certain. Therefore, credit protection still provides him with an incentive to hold out. 181 One possibility, however, is that the sovereign’s default on any post-restructuring residual debt might trigger the swap. However, this possibility is insufficient to convince the protected holder to participate, for at least two reasons. First, there is no guarantee that there will be any residual debt, or that the debtor will default on it. The only means the protected holder possesses to ensure that a post-restructuring default will occur (or that she will retain the full benefit of the old bonds) is to hold out and keep the old bonds herself. Second, even if a post-restructuring default occurs, the amount of the residual debt might not be sufficient for the default to meet the Payment Requirement and trigger the swap. Of course, there is always the possibility that the voluntary exchange might fail sufficiently to improve the country’s financial position to avoid default on its entire debt. This is what happened in Argentina, as explained above. Thus, most swaps were eventually triggered on December 21. The issue settled in Eternity arose with respect to credit default swaps which matured between November 1st and December 21st.

- 68 -

and will fail, thus preventing the sovereign from successfully restructuring its bonds and

reducing its debt load. Notably, this is the case regardless of how generous the exchange

offer is to bondholders, since any reduction in the bond’s value will suffice to provide

protected holders with an incentive to hold out. This argument also assumes that the returns

on the various enforcement techniques available to creditors are low. Therefore, it would

survive a rectification of the Elliott interpretation of the pari passu clause. As a result, the

existence of credit protection increases the likelihood of failure for voluntary exchange offers.

Moreover, even if the exchange succeeds despite the non-participation of protected

bondholders, their actions still increase the likelihood of post-restructuring holdout litigation.

This is because, after the exchange offer is consummated and followed by a default on the

residual debt, the protection sellers end up with the defaulted obligations. While participating

in the exchange offer would have been the rational decision ex ante, they no longer have that

option. Instead, they will likely choose to take their chances in court, or to cut their losses by

selling the obligations at a discount to vulture funds who will attempt to sue the debtor. In

both cases, the chances of recovery are uncertain, but the creditors might at the very least

succeed in disrupting the debtor’s international trade and financial operations. Thus, in

addition to increasing the risk of failure for voluntary exchange offers, the presence of

protected holders also aggravates the risk that a successful exchange will be followed by

disruptive holdout litigation.

- 69 -

4. Amendments to the ISDA Definitions

As most sovereign credit derivatives are still documented under the 1999

Definitions,182 the incentives problem they cause can be expected to persist for some time.

Nevertheless, although the fragility of the current restructuring process itself is at the heart of

the problem, changes in documentation and market practices could alleviate its effects.

As mentioned above, the Restructuring clause was modified after Conseco to exclude

renegotiations of small classes of obligations and restructurings effected with low majorities,

as well as to impose a maturity limitation date for deliverable obligations. The modified

clause, however, is not usually included in sovereign CDSs.183 In addition, even if it were so

included, it would not in itself solve the problem, as all it does is further narrow the

Restructuring definition. If the main Restructuring definition did encompass voluntary

sovereign restructurings effected through exchanges, then the modifications could be

appropriate to exclude certain types of opportunistic restructurings. In their present form,

however, the post-Conseco amendments have little impact on sovereign finance.

The 2003 Restructuring definition, for its part, eliminates both the concept of

Obligation Exchange and the requirement that an exchange be ‘mandatory.’ Instead,

Restructuring only applies when one of the enumerated events (i) occurs “in a form that binds

all holders of such obligations,” (ii) “is agreed between the Reference Entity or a

Governmental Authority and a sufficient number of holders of such Obligation to bind all

holders of the Obligation,” or (iii) “is announced (or otherwise decreed) by a Reference Entity

182 See Packer and Suthiphongchaim, supra note 5, at 83. 183 See Id.

- 70 -

or a Governmental Authority in a form that binds all holders of such Obligations.” 184

However, since without a collective action clause, the changes that trigger a Restructuring

cannot be imposed on an unwilling minority of creditors, none of the alternatives applies to

“voluntary” sovereign debt restructurings.

On one possible interpretation, alternative (i) or (ii) would be met if all the holders in

fact consented to the restructuring by tendering their obligations in response to an exchange

offer. Thus, a voluntary debt exchange would not trigger a Restructuring Credit Event unless

and until all holders tendered their obligations. This interpretation of clause (i), however, is

unlikely to prevail. First, as a practical matter, some holders of sovereign bonds virtually

always omit to participate in exchanges for a variety of reasons, so that even under a de facto

test, clause (i) would usually not be satisfied. Second, the definition seems on its face to

contemplate de jure, rather than de facto, bindingness. Indeed, if all creditors individually

agreed to a restructuring, it could be argued that it does not have any of the coercive elements

that normally accompany a credit event.

For these reasons, the 2003 Restructuring clause will likely be interpreted along the

same lines as the 1999 one, with the result that voluntary exchanges will not trigger swaps

documented under the new definitions. The exclusion of voluntary exchanges from the

Restructuring definition reflects market practice, and many participants are satisfied with

Eternity’s treatment of the Argentine exchange. One possible reason is the difficulty of

distinguishing between truly distressed exchanges and non-coercive, routine liability

management operations which include an exchange offer. Of course, even if all voluntary

restructurings were prima facie caught by paragraph (a) of the definition, the benign sort 184 See Appendix 1 for the full definition.

- 71 -

would be excluded by paragraph (b)(iii), as they do not reflect a deterioration of the

sovereign’s creditworthiness. That clause, however, is fundamentally subjective. Frequent

resort to it would significantly increase legal uncertainty and slow down settlement. Another

reason for the market’s preference for excluding voluntary restructurings is that if they were

credit events, protection buyers might have reduced incentives to negotiate the best possible

arrangement. After the restructuring, protection sellers would be left with excessively

devalued obligations. On the other hand, as restructuring negotiations are usually conducted

by large, unprotected creditors, this problem might not arise in practice as small protected

creditors would free-ride on institutions-led restructuring negotiations.185

As a result, even as the sovereign credit derivatives market gradually migrates to the

2003 ISDA Definitions, the distortion in restructuring incentives created by these instruments

will endure.

C. Collective Action Clauses

Would the use of CACs to effect a sovereign debt restructuring trigger credit default

swaps referenced to the sovereign? Once again, the issue only arises with respect to pre-

default restructurings, since the swaps would already have been triggered by a prior default.

1. Collective Action Clauses and the ISDA Definitions

As mentioned above, the changes in financial terms adopted as part of a restructuring

normally fit one or more of the events listed in the Restructuring definition. The 1999

Definition covers such events when they are “agreed between the Reference Entity … and the

185 Another possible solution to this problem would be to give protection sellers a greater role in negotiations with distressed debtors: see Pollack, supra note 107, at 47ff. However, Ms. Pollack also indicates that this would go against the prevailing market sentiment that credit default swaps should remain “separate and independent” products. Id. at 51.

- 72 -

holder or holders of such Obligation.”186 The 2003 Definition contemplates CACs even more

clearly, by providing that changes to the listed financial terms constitute a Restructuring if

they “are agreed between the Reference Entity … and a sufficient number of holders of such

Obligation to bind all holders of the Obligation.”187 Thus, in both cases, restructurings

effected through CACs are prima facie caught by the Restructuring definition.

Both definitions, however, go on to exclude such events if they are “provided for

under the terms of [the] Obligation in effect as of the later of the Trade Date and the date as of

which such Obligation is issued or incurred.” 188 Clearly, the CAC is part of the debt

instrument itself, so that implementation of a restructuring through its use might appear to be

“provided for” under the terms of the Obligation. On this interpretation, no Restructuring

credit event would be triggered. There are, however, multiple reasons why such an

interpretation ought to be rejected.

First, the drafters of the ISDA definition clearly did not intend the “provided for”

clause to exclude CACs. To be sure, the 1999 Restructuring clause does not expressly

describe CACs, and one might argue that the original drafters did not turn their minds to this

problem and had no specific intention to cover or exclude CACs. After all, the inclusion of

CACs in corporate bonds is prohibited by the Trust Indenture Act.189 Likewise, until very

recently, sovereign bonds issued in the United States did not include CACs. 190 This

argument, however, overlooks the fact that the Definitions are meant to be used to document

transactions under either New York or English law. Bonds issued in London, both corporate 186 1999 DEFINITIONS, s. 4.7(a) (see Appendix 1). 187 2003 DEFINITIONS, s. 4.7(a) (see Appendix 1). 188 1999 DEFINITIONS, s. 4.7(a); the 2003 Definitions contain a substantially identical clause, but use the expression “expressly provided.” 189 Supra note 67 . 190 See supra, Part I.

- 73 -

and sovereign, routinely include CACs, and these provisions are frequently used to implement

restructurings. In light of the drafters’ knowledge of this fact, their failure clearly to exclude

CACs can only be read as intention that their use trigger the Restructuring clause.

This result is even clearer under the 2003 Definitions. Clause (ii) of the 2003

Definition so clearly covers CACs that it is almost inconceivable that the drafters nevertheless

wanted to exclude them. Another indication of this intent is that, in addition to modifying the

Restructuring clause to cover CACs, the 2003 drafters also added provisions that make

restructured sovereign obligations deliverable despite changes in their characteristics caused

by the restructuring.191 This clearly indicates that they contemplated that, at least in certain

circumstances, sovereign debt restructurings could trigger the Restructuring event. As seen

above, voluntary exchanges do not trigger the clause, so the drafters must have been

contemplating the use of CACs.

Second, any interpretation that would exclude CACs from the Restructuring definition

would clearly be at odds with the policy pursued by the adoption of CACs. These clauses

allow a supermajority of holders to bind the minority to the financial terms of a restructuring.

At least with respect to this feature, they are intended to solve the collective action problem

that otherwise makes restructurings difficult or impossible, and to eliminate the holdout

creditor problem. Their adoption is seen as a market-based alternative to a more elaborate

international bankruptcy regime.192 Thus, their use, as contemplated by supporters, can be

expected to reflect a substantial deterioration in the sovereign’s creditworthiness. This is 191 Section 2.16 of the 2003 Definitions defines “Sovereign Restructured Deliverable Obligation” as any sovereign obligation that has all the Deliverable Obligation Characteristics (and is thus deliverable) immediately prior to a restructuring, regardless of whether it still has all these characteristics after the restructuring. Section 2.15(c) provides that a Sovereign Restructured Deliverable Obligation is deliverable in a sovereign restructuring, subject to certain conditions. 192 See, e.g., Taylor, supra note 72.

- 74 -

admittedly equally true of voluntary restructurings under the current system, which do not

trigger credit events. However, when a CAC is used, the protection buyer’s argument that the

exchange is in fact involuntary is stronger, because the restructuring can be enforced against

him even if he votes against it. Indeed, even if a protection buyer votes for a restructuring,

some coercion still occurs, as the negotiations are conducted under the shadow of the CAC’s

potential utilization.

Finally, excluding collective action clauses from the definition would perpetuate the

incentives problem created by credit derivatives in the restructuring process. CAC have not

yet been used in actual restructurings and, as noted above, several commentators doubt their

usefulness. An interpretation of the definition that would exclude CACs would further dilute

their effect. Like protected holders faced with a voluntary exchange offer, protected holder

faced with a CAC restructuring will prefer to hold out in hope that the restructuring will fail

and that the debtor will default, thus triggering the swaps. While the problem would be

somewhat alleviated by the lower voting threshold under CACs – 75% or 85%, compared to

the de facto 90% threshold for a voluntary exchange – a small number of protected creditors

could still tip the scales.

What, then, is one to make of the “provided for” clause? As a starting point, one

should note that the clause only applies when the relevant “event” is not provided for under

the terms of the obligation. The term “event” refers to the list that follows the first paragraph

of Section 4.7(a): reduction in interest or principal, extension of maturity, change in currency,

etc. CACs in and of themselves, however, provide for no such “event;” all they do is create

an amendment procedure that allows a majority of holders to amend the obligations.

Interestingly, while the “provided for” clause predated by several years the Restructuring

- 75 -

definition debate outlined above, as well as ISDA’s effort to address sovereign restructurings

in the 2003 Definitions, it survived these changes. The 2003 drafters, however, qualified it by

adding the word “expressly,” which reinforces the conclusion that they intended the clause to

cover express provisions rather than amendments later made to the obligations. An example

of an express provision would be a clause that automatically adjusts the interest rate on the

basis of some exogenous event, such as a credit rating improvement. While this adjustment is

clearly excluded by the “provided for” clause, use of CACs to restructure the obligations is

not.

2. Do Collective Action Clauses Make the Obligations Contingent?

In addition to the “provided for” clause, another difficulty arises with respect to CACs.

One of the Deliverable Obligations Characteristics often specified by parties to sovereign

CDSs is “Not Contingent.” 193 This means, as the term suggests, that in order for the

protection buyer to be entitled to deliver an obligation under a physically-settled swap, that

obligation must not be subject to “any contingency.”194 This clause was litigated in British

courts in 2003, in the aftermath of Railtrack’s bankruptcy. Nomura International plc had

purchased credit default swaps referenced on Railtrack from Credit Suisse First Boston.

When the credit event occurred, Nomura attempted to deliver Railtrack exchangeable bonds

to Credit Suisse. These bonds could, under certain circumstances, be exchanged for Railtrack

193 As Ranciere, supra note 110, indicates at 17, “[i]n emerging markets, physical settlement tends to be the dominant procedure,” for several reasons: (i) accurate quotes are difficult to obtain for distressed sovereign bonds during a financial crisis, as the market is highly illiquid; (ii) in such circumstances, the recovery value of bonds is often underestimated, so that cash settlement would expose the protection seller to greater risk; and (iii) protection buyers often use sovereign CDSs to hedge actual positions. Although there is no market data on the use of the “Non Contingent” characteristic, it appears to be a standard term. For instance, the 2003 ISDA Sovereign Master Credit Derivatives Confirmation Agreement, a model published to facilitate the documentation of sovereign credit derivatives under the 2003 Definitions, recommends the inclusion of “Non Contingent” for all such derivatives. 194 1999 DEFINITIONS, 2.18(b)(vii)

- 76 -

shares upon demand by the holder or the Trustee acting for the benefit of holders. Railtrack

did not have the option to demand that the bonds be exchanged. Despite this, Credit Suisse

refused to settle the swap and argued that the exchange feature made the bonds “contingent”

and thus non-deliverable.

In Nomura International plc v. Credit Suisse First Boston International,195 Justice

Langley, of the Commercial Division, held that Nomura was entitled to deliver the bonds. In

his view, a provision that operates in favor of a bondholder and in his favor could not be

described as a contingency. Despite the exchangeability feature, the holder can choose to

claim repayment and is entitled to it, as the exchange is within his control and is unaffected by

external events or by the debtor’s actions. Langley J. summarized his position as follows:

The purpose of requiring that the Deliverable Obligation be “Not Contingent” is, I think, plainly to secure a right to payment of the principal amount as such which cannot be affected in amount by extraneous factors over which the seller of credit, as holder, has no control. [emphasis added]

This language raises an issue with respect to bonds containing CACs. On a literal

reading of Justice Langley’s decision, since other holders of the bonds can agree with the

debtor to change the bonds’ financial terms and impose the new terms on a dissenting

minority, an individual holder’s right to payment can be affected by “extraneous factors” over

which he has no control. Therefore, the bonds are “contingent” and therefore non-deliverable

under the terms of most sovereign CDSs currently on the market. Notably, under this literal

interpretation, clauses permitting majority amendments to non-payment terms would not

make obligations contingent, since they do not affect their “amount.” Likewise, clauses

permitting unanimous amendments to payment terms would not be affected: since every

195 2003 WL 933569 (QBD (Comm. Ct.)). On this dispute, see Pollack, supra note 107 at 12-15.

- 77 -

individual holder has a veto, the amendments are never beyond his “control.” Only collective

action clauses, which allow majority amendments to payment terms, would make bonds

contingent.

This interpretation, of course, would be highly problematic. CDSs written on

sovereigns that have converted their existing bonds to new ones containing CACs would be

useless. Even more strikingly, all English law bonds containing CACs would be non-

deliverable. This result would clearly defeat the legitimate expectations of protection buyers.

Of course, this problem could be avoided by omitting to list “Not Contingent” as a

Deliverable Obligation Characteristic in CDSs on sovereigns who include CACs in their

bonds. This, however, might jeopardize the protection sellers’ interest in excluding

“genuinely” contingent obligations. ISDA recognized this in 2003, as it amended the

Definitions to clarify that exchangeable and convertible obligations are not contingent. The

new language, however, does not clearly address the potential argument that CACs make

obligations contingent.

The better view is that voting provisions such as CACs are simply not the kinds of

contingency that should make bonds non-deliverable. As explained above, they are intended

as a substitute for a statutory restructuring process that is absent from sovereign finance. The

possibility that such procedures of general application might be brought to bear on a debtor

and lead to a reduction of its creditor’s claims is a normal background expectation in all

financial markets. In other words, just as the possibility that a corporate debtor may go

bankrupt does not make its obligations “contingent,” neither does the possibility that a

sovereign debtor may invoke voting procedures that have been put in place to facilitate the

restructuring of distressed debt. One should not, however, overlook the potential for majority

- 78 -

abuse that comes with CACs. In a figurative sense, one might say that an individual

creditor’s claim is always “contingent” on the other creditors’ not colluding with the debtor to

reduce its value. It may be that the solution to this problem should come from the courts’

imposing some duty of fair dealing on the majority, as advocated by Professors Bratton and

Gulati.196

D. The SDRM

None of the more ambitious sovereign bankruptcy proposals is currently attracting

significant support in policy circles. It is, however, worthwhile to examine the interaction of

a potential Sovereign Debt Restructuring Mechanism (SDRM) with credit derivatives, as the

reform project may be revived if CACs prove insufficient to facilitate debt restructurings in

the future.

Should such a far-reaching reform of sovereign debt restructuring be adopted, one

expects that ISDA would change its definitions, perhaps in substantial ways, to reflect the

new environment. Even under the current definitions, however, a sovereign debtor’s decision

to invoke the SDRM would constitute a credit event and trigger CDSs written on that

sovereign. First, use of the SDRM would undoubtedly trigger the Bankruptcy definition.

This credit event is not usually included in sovereign CDSs, but market practice would likely

evolve in that direction following the adoption of the SDRM. Second, the Failure to Pay and

Repudiation/Moratorium credit events would also apply to the payments standstill following

the debtor’s filing. Finally, physical settlement of the swaps would not be subject to any

196 See Bratton & Gulati, supra note 78.

- 79 -

contingency problem created by CACs, since the SDRM would be provided for by statute or

by international treaty rather than in the obligations themselves.

This result is consistent with the parties’ expectations, since use of the SDRM is

clearly meant to be a last resort for sovereigns with unsustainable debt. Under the Krueger

proposal, the IMF or an independent body would have to confirm that the sovereign’s

financial position justifies invoking the SDRM, thus providing some protection against

unjustified restructurings.197 Therefore, use of the SDRM would in virtually all cases reflect a

significant deterioration in the sovereign’s creditworthiness, and thus a materialization of the

sort of credit risk the swaps are intended to cover. The standstill and the subsequent reduction

or rescheduling of payments on the sovereign’s debt would also have coercive features

strongly analogous with those of domestic bankruptcy regimes.

Once again, however, one might wonder whether the incentives created by the

mechanism are truly optimal. Dr. Krueger, for instance, argues that the SDRM would provide

a legal background for voluntary restructuring negotiations. On that view, future

restructurings, like arrangements with creditors in the domestic context, would be negotiated

“in the shadow of the law,” in the sense that all involved parties would know that the likely

alternative to an agreement is for the sovereign to invoke the SDRM. In Dr. Krueger’s words,

“the intention is that the existence of a predictable legal mechanism will in itself help debtors

and creditors to reach agreement without the need for formal activation.”198 This benefit of a

SDRM, however, would be undermined by the fact that while a voluntary restructuring does

not trigger CDSs, use of the SDRM does. Therefore, protected holders would have incentives

197 See A NEW APPROACH, supra note 6, at 23-28. 198 Id. at 4.

- 80 -

to refuse a negotiated settlement and force the debtor to use the SDRM. While this result is

preferable to an outright default, it nevertheless mitigates the expected benefits of the SDRM.

As a result, even if the SDRM were to be adopted, the exclusion of voluntary exchanges from

the Restructuring definition would hinder the resolution of sovereign debt problems.

E. Sovereign Debt and Global Financial Stability

In addition to the legal interpretation issues and incentives problems discussed above,

the potential effects of the rise of sovereign credit derivatives on financial stability deserve

mention.

1. Market Concentration and Systemic Risk

The structure of the credit derivatives market raises systemic risk concerns.

Intermediation activities with respect to credit derivatives are highly concentrated,199 often in

unregulated or lightly regulated subsidiaries of large securities firms. A small number of

dealers routinely buy and sell credit derivatives, either on their own account or as

intermediaries. Among other activities, they maintain a substantial inventory of such

derivatives for the purpose of future business with parties wishing to take opposite positions.

Although these dealers are thought to maintain a balanced book, verifying this is

difficult under the current regulatory regime.200 Moreover, the considerable legal uncertainty

surrounding credit derivatives means that even an apparently balanced book may turn out to

be questionable under pressure.201 For instance, dealers who acted as intermediaries between

199 See CREDIT RISK TRANSFER, supra note 5, at 27. 200 See id. at 25-26; see also Garry J. Schinasi et al., Modern Banking and OTC Derivatives Markets: The Transformation of Global Finance and its Implications for Systemic Risk, IMF Occasional Paper 203 (2000), at 49ff. For an examination of the credit requirements and other regulations applicable to derivatives dealers, see INTERNATIONAL FINANCE, supra note 10, Draft ch. 15, at 22-45. 201 See id. at 27.

- 81 -

parties to Argentine swaps have been known to take contrary positions in litigation arising

under each branch of the transaction. 202 Inconsistent court rulings and the resulting

imbalances in payments could threaten the liquidity of some dealers. Likewise, the difficulty

of hedging the exposures created by CDSs by purchasing identical instruments can also lead

to mismatches between the losses on credit derivatives and the payments received under

other, less accurate hedging techniques, such as shorting the underlying obligations

themselves.203

These systemic risk concerns are compounded by the fact that, in addition to dealer

concentration, the sovereign derivatives market is also highly concentrated on some reference

entities. As seen above, a small number of major emerging debtors, including Brazil, Mexico,

Japan, the Philippines and South Africa, account for a vast proportion of the market.204 While

the concept and mechanics of financial crisis ‘contagion’ in emerging markets are still a

matter of debate among economists, there is no doubt that series such crises have happened in

quick succession in the past. One needs only think of the Latin American crisis of 1982 and

the East Asian crisis, Russian default and Brazilian scare of 1997-98. A similar series of

closely correlated crises and the resulting credit events could severely strain the credit

derivatives market.205 The execution of all requested settlements would require that the

system operate without a glitch, and a failure by a dealer or another party to many transactions

could easily spread to others.

202 See Pollack, supra note 107, at p. 19-20. 203 On this point, see Ranciere, supra note 110, at 19. 204 See Packer and Suthiphongchai, supra note 5, at 81-83; Ranciere, supra note 110, at 5. 205 On the financial stability problems associated with credit derivatives exposures to correlated credit risks, see generally David Rule, The Credit Derivatives Market: Its Development and Possible Implications for Financial Stability, FIN. STABILITY REV., June 2001, at 117.

- 82 -

Evaluating the probability of such a series of events, and whether it would suffice to

trigger a series of failures in the international financial system, is beyond the scope of this

paper. It is certainly encouraging that, so far, credit derivatives have performed satisfactorily

in periods of financial stress. Protection buyers have been able to recover substantial amounts

from their investments in the Industrial Finance Corporation of Thailand and the Korean

Development Bank.206 Likewise, despite the initial confusion caused by Argentina’s ill-fated

November 2001 voluntary exchange, credit default swaps were orderly settled when the

country actually defaulted.207 In Russia, however, the 1998 default led to a number of legal

challenges relating to credit derivatives, including the ANZ Banking case. As the market

keeps expanding, the possibility of a chain reaction remains worrisome.

Finally, credit derivatives also reduce systemic risk by spreading exposures among

financial institutions and investors and by facilitating diversification. Moreover, the many

technical amendments to the 2003 Definitions are likely to improve legal certainty in the

credit derivatives market. Therefore, it may be that, given a sufficiently robust legal regime,

sovereign credit derivatives actually result in a net decrease in overall systemic risk.

2. Excessive Lending in Emerging Markets

The availability of credit protection may contribute to continued excessive lending in

emerging markets. For instance, large financial institutions may be induced to make riskier

loans and increase their sovereign debt holdings by the confidence that the credit risk can be

passed on to other institutions such as smaller banks or insurance companies. Sophistication

about the characteristics and risks associated with derivatives is spreading rapidly through the 206 See André Scheerer, Credit Derivatives: An Overview of Regulatory Initiatives in the U.S. and Europe, 5 FORDHAM J. CORP. & FIN. L. 149, 152-53 (2000). 207 See Ranciere, supra note 110, at 22-23.

- 83 -

financial system. In many cases, these instruments offer smaller institutions opportunities for

diversification that would otherwise be unavailable. Nevertheless, allegations intermittently

surface that complex derivatives sold by large dealers have caused substantial losses to

relatively unsophisticated buyers who failed fully to appreciate their risks.208 While this

element of information asymmetry may subside over time, many commentators believe that

financial markets as a whole have repeatedly shown excessive enthusiasm for emerging

markets,209 which could be further fueled by the multiplication and greater accessibility of

synthetic exposures created by credit derivatives.

CONCLUSION

This examination of the contractual framework for sovereign credit derivatives invites

three general conclusions. First, the 2003 Definitions evidence a conscious effort by ISDA to

respond systematically to the legal issues raised by sovereign restructurings. This is a

welcome change from the Association’s previous approach of responding to discrete market

events by ad hoc amendments to the Definitions. As a result, the 2003 Definitions, when

interpreted correctly, provide clear guidance with respect to the various forms a sovereign

restructuring may take, including the utilization of the newly-adopted CACs in New York

sovereign bonds. As the credit derivatives market moves to the new definitions, legal

uncertainty will be reduced.

208 In the mid-1990s, considerable litigation arose in the United States alleging improprieties by dealers in marketing derivatives to end-users. See INTERNATIONAL FINANCE, supra note 10, Draft ch. 15, at 46-58. The IAIS PAPER, supra note 117, at 5-6, notes that credit risk transfer activities pose particular challenges for insurance companies’ risk management programs, and that regulators should take measures to ensure that the firms under their supervision have the requisite skills and knowledge to evaluate and manage the risks involved. 209 See, e.g., STIGLITZ, supra note 61.

- 84 -

As a more general matter, however, the tension between the market’s apparent

satisfaction with the Eternity ruling, on the one hand, and its undesirable consequences on the

feasibility of sovereign debt restructurings, on the other, points to the externalities associated

with private rulemaking with respect to derivatives. In other words, even assuming that

ISDA, taking into account the interests of both protection buyers and sellers, can elaborate

definitions that optimally facilitate private risk management, the danger remains that such

definitions may be suboptimal from the perspective of global financial stability.

The extent to which this tension is real remains unclear. As explained above, the

exclusion of voluntary exchanges from the Restructuring definition was a deliberate choice on

the part of the drafters, especially in 2003, when sovereign restructuring issues were clearly

considered. It might be that the fall of regulatory obstacles will lead ISDA to comply with

market demands and abandon the Restructuring definition. ISDA would then have to

articulate an alternative definition to address sovereign debt restructurings. In view of the fact

that many of the techniques used to reduce moral hazard in corporate credit derivatives are not

available with respect to sovereigns, protection sellers focusing on the problem might demand

that voluntary restructurings be included in one form or another. One possible option would

be to defer to the judgment of some third party, such as a credit rating agency, as to whether

the exchange is ‘distressed’ or not. This would bring ISDA documentation in line with

financial stability imperatives. They would also make sense because, as indicated above, the

absence of a bankruptcy regime for countries creates substantial differences between the

- 85 -

sovereign and domestic restructuring processes, with the result that what is truly “voluntary”

in one case should not be treated the same in the other.210

This solution, however, seems improbable given ISDA’s choice to retain the 1999

definitional approach to Restructuring in its 2003 definitions, despite the objections of some

market participants. Moreover, changes to the definitions cannot in and of themselves address

the fundamental issue at hand. The real threat to stability does not come from credit

derivatives themselves, but from the fragility of the current restructuring process. Progress in

avoiding financial crises can only be achieved by improvements to this process. For instance,

as the sovereign debt market migrates to CACs, the incentives problems created by credit

protection are likely to decrease.

Second, while the 2003 ISDA drafters were clearly aware of the restructuring debate,

international financial policymakers do not appear to have given much though to the impact

of credit derivatives on their reform plans. This is unsurprising under the normal expectation,

which is that financial markets will spontaneously adapt to major structural changes. This

expectation, however, can become rather perilous when the interests of the private parties

drafting the clauses may be at odds with the stability objectives pursued by the reforms. More

generally, the lesson is that, given the growing importance of financial derivatives, their

interaction with any proposed innovation ought to be closely scrutinized in order to avoid

unintended consequences. This point, of course, applies not only to sovereign restructuring

reform but to almost any serious financial reform effort.

210 For instance, Moody’s makes the point that “[c]redit risk is risk imposed on the lenders by an obligor. It does not include, and a Moody’s rating does not address, risks that lenders impose on themselves – i.e., the possibility that lenders will voluntarily take losses that are not forced on them by obligors.” (Tolk, supra note 116, at 8). It is unclear, however, whether “voluntary” sovereign restructurings truly fit this description.

- 86 -

Finally, at the intersection of sovereign debt and credit derivatives, the problems one

observes are the same as in these individual markets, but they compound each other to a

considerable degree. The brittleness of the current restructuring process makes it vulnerable

to the destabilizing effect of credit derivatives. In turn, the legal uncertainties associated with

credit risk transfer activities may amplify the systemic risk created by the flow of capital to

emerging markets. In order for this rapidly expanding market to confer all its benefits without

endangering global financial stability, these issues will need to be explored empirically.

- 87 -

APPENDIX: ISDA RESTRUCTURING DEFINITIONS

1999 ISDA Definitions: Section 4.7. Restructuring. (a) “Restructuring” means that, with respect to one or more Obligations, including as a result of an Obligation Exchange, and in relation to an aggregate amount of not less than the Default Requirement, any one or more of the following events occurs, is agreed between the Reference Entity or a Governmental Authority and the holder or holders of such Obligation or is announced (or otherwise decreed) by a Reference Entity or a Governmental Authority in a form that is binding upon a Reference Entity, and such event is not provided for under the terms of such Obligation in effect as of the later of the Trade Date and the date as of which such obligation is issued or incurred:

(i) a reduction in the rate or amount of interest payable or the amount of scheduled interest accruals; (ii) a reduction in the amount of principal or premium payable at maturity or at scheduled redemption dates; (iii) a postponement or other deferral of a date or dates for either (A) the payment or accrual of interest or (B) the payment of principal or premium;

(iv) a change in the ranking in priority of payment of any Obligation, causing the Subordination of such Obligation; or (v) any change in the currency or composition of any payment of interest or principal. […]

(b) Notwithstanding the provisions of Section 4.7(a), none of the following shall constitute a Restructuring: […]

(iii) the occurrence of, agreement to or announcement of any of the events described in Section 4.7(a)(i) to (v) in circumstances where such event does not directly or indirectly result from a deterioration in the creditworthiness or financial condition of the Reference Entity.

(c) If an Obligation Exchange has occurred, the determination as to whether one of the events described under Section 4.7(a)(i) to (v) has occurred will be based on a comparison of the terms of the Obligation immediately before such Obligation Exchange and the terms of the resulting Obligation immediately following such Obligation Exchange. […] Section 4.9. Obligation Exchange. “Obligation Exchange” means the mandatory transfer (other than in accordance with the terms in effect as of the later of the Trade Date or date of

- 88 -

issuance of the relevant Obligation) of any securities, obligations or assets to holders of Obligations in exchange for such Obligations. When so transferred, such securities, obligations or assets will be deemed to be Obligations. 2003 ISDA Definitions: Section 4.7. Restructuring. (a) “Restructuring” means that, with respect to one or more Obligations and in relation to an aggregate amount of not less that the Default Requirement, any one or more of the following events occurs in a form that binds all holders of such Obligation, is agreed between the Reference Entity or a Governmental Authority and a sufficient number of holders of such Obligation to bind all holders of the Obligation or is announced (or otherwise decreed) by a Reference Entity or a Governmental Authority in a form that binds all holders of such Obligations, and such event is not expressly provided for under the terms of such Obligation in effect as of the later of the Trade Date and the date as of which such Obligation is issued or incurred:

(i) a reduction in the rate or amount of interest payable or the amount of scheduled interest accruals; (ii) a reduction in the amount of principal or premium payable at maturity or at scheduled redemption dates; (iii) a postponement or other deferral of a date or dates for either (A) the payment or accrual of interest or (B) the payment of principal or premium; (iv) a change in the ranking in priority of payment of any Obligation, causing the Subordination of such Obligation to any other Obligation; or (v) any change in the currency or composition of any payment of interest or principal to any currency which is not a permitted currency. […]

(b) Notwithstanding the provisions of Section 4.7(a), none of the following shall constitute a Restructuring: […]

(iii) the occurrence of, agreement to or announcement of any of the events described in Section 4.7(a)(i) to (v) in circumstances where such event does not directly or indirectly result from a deterioration in the creditworthiness or financial condition of the Reference Entity.

- 89 -

BIBLIOGRAPHY

Primary Sources and Official Documents BRITISH BANKERS’ ASSOCIATION (BBA), CREDIT DERIVATIVES REPORT 2001/2002 (2002). COMMITTEE ON THE GLOBAL FINANCIAL SYSTEM (CGFS), CREDIT RISK TRANSFER (2003) Federative Republic of Brazil, 10% Global Bonds Due 2007, Prospectus Supplement Dated April 29, 2003 FINANCIAL SERVICES AUTHORITY, CROSS-SECTOR RISK TRANSFERS (2002) Foreign Sovereign Immunities Act of 1976, Pub. L. No. 94-583, 90 Stat. 2891 (codified at 28 U.S.C. §§1330, 1602-11 (2000)). Group of Seven, Strenghtening the International Financial System and the Multilateral Development Banks, Report of G-7 Finance Ministers and Central Bank Governors, available at http://www.g8.utoronto.ca/finance/fm010707.htm (July 7, 2001). INTERNATIONAL ASSOCIATION OF INSURANCE SUPERVISORS, IAIS PAPER ON CREDIT RISK TRANSFER BETWEEN INSURANCE, BANKING AND OTHER FINANCIAL SECTORS PRESENTED TO THE FINANCIAL STABILITY FORUM (2003) International Monetary Fund, Progress Report to the International Monetary and Financial Committee on Crisis Resolution (Sept. 5, 2003), available at http://www.imf.org/external /np/pdr/cr/2003/eng/090503.pdf INTERNATIONAL SWAPS AND DERIVATIVES ASSOCIATION (ISDA), 1999 ISDA CREDIT DERIVATIVES DEFINITIONS (1999) ISDA, 2002 MARKET SURVEY: HISTORICAL DATA (2003) ISDA, 2002 MASTER AGREEMENT (2002) ISDA, 2003 ISDA CREDIT DERIVATIVES DEFINITIONS (2003) ISDA, USER’S GUIDE TO THE ISDA 2002 MASTER AGREEMENT (2003) KRUEGER, ANNE O., A NEW APPROACH TO SOVEREIGN DEBT RESTRUCTURING (2002) Republic of South Africa, 5.25% Notes Due May 16, 2013, Prospectus Supplement Dated May 9, 2003

- 90 -

República Oriental del Uruguay, Exchange Offer, Prospectus Supplement Dated April 10, 2003 State Immunity Act, 1978, 26 Eliz. II, 33 (England), 17 I.L.M. 1123. Trust Indenture Act of 1939, 15 U.S.C. ch. 2A United Mexican States, 6.625% Global Notes Due 2015, Prospectus Supplement Dated February 26, 2003 United States Bankruptcy Code, 11 U.S.C. 101ff. Cases Allied Bank International v. Banco Credito Agricola de Cartago, 757 F.2d 516 (2d Cir. 1985), cert. dismissed, 473 U.S. 934 (1985) CIBC Bank & Trust Co. (Cayman) Ltd. v. Banco Central do Brazil, 886 F.Supp. 1105 (S.D.N.Y. 1995) Deutsche Bank v. ANZ Banking Group, 2000 WL 1151384 (QBC (Comm. Ct.)) Elliott Associates v. Republic of Peru, 194 F.3d 363 (2d Cir. 1999) Elliott Assocs., L.P., General Docket No. 2000/QR/92 (Court of Appeals of Brussels, 8th Chamber, Sept. 26, 2000) Eternity Global Master Fund v. Morgan Guaranty Trust Co. of New York, 2002 WL 31426310 (S.D.N.Y.), and 2003 WL 21305355 (S.D.N.Y.) Libra Bank Ltd. v. Banco National de Costa Rica, 570 F.Supp. 870 (S.D.N.Y. 1983) Macrotecnic Int’l Corp. v. Republic of Argentina, Statement of Interest of the United States, 02 CV 5932 (TPG), January 12, 2004 Nomura International v. Credit Suisse First Boston International, 2003 WL 933569 (QBD (Comm. Ct.) Pravin Banker Associates v. Banco Popular del Peru, 109 F.3d 850 (2d Cir. 1997) Ursa Minor v. Aon Financial Products, 7 Fed.Appx. 129 (2d Cir. 2001) Weltover v. Argentina, 504 U.S. 607 (1992) Books and Periodicals

- 91 -

Akerlof, George, The Market for Lemons: Quality Uncertainty and the Market Mechanism, 84 Q.J. ECON. 488 (1970) Alden, Edward, Catan, Thomas & Lapper, Richard, Argentina’s President Quits; At Least 14 Killed in Riots as Country Falls into Chaos; Leadership Battle Breaks Out, FIN. TIMES, Dec. 21, 2001, at 1 Anonymous, Nestor Kirchner's nimble cookery – Argentina’s default, and its deal with the IMF, ECONOMIST, Sept. 13, 2003 Anonymous, SDRM Is Dead, and That’s Official, EUROMONEY, May 2003 Asiedu-Akrofi, Derek, A Comparative Analysis of Debt-Equity Swap Programs in Five Major Debtor Countries, 12 HASTINGS INT’L & COMP.L.REV. 537 (1989) Authors, John, Mexico Pioneers a Plan to Ease Debt, FIN. TIMES, Feb. 25, 2003 Authors, John, Mexico Sends Signal with Bond Clauses, FIN. TIMES, Feb. 27, 2003, at 23 Batchelor, Charles, Credit Default Swaps Join Booming Derivatives Line-Up, FIN. TIMES, Feb. 11, 2004, at 26 Beattie, Alan, Financial Grouping Wants New Debt Rules, FIN. TIMES, June 12, 2002; SDRM Finds Few Friends in the Markets, EUROMONEY, Nov. 2002 Beattie, Alan, Uruguay provides test case for merits of voluntary debt exchange, FIN. TIMES, April 23, 2003, at 3 Bell, Ian & Dawson, Petrina, Synthetic Securitization: Use of Derivative Technology for Credit Transfer, 12 DUKE J. INT’L & COMP. L. 541 (2002) Bendernagel, Don, et al., Credit Derivatives: Usage, Practice and Issues, in SWAPS AND OTHER DERIVATIVES IN 2001 (Edward J. Rosen ed., 2001) BLUSTEIN, PAUL, THE CHASTENING (2003) Bratton, William W. & Gulati, G. Mitu, Sovereign Debt Restructuring and the Best Interests of Creditors, 57 VAND. L. REV. (forthcoming January 2004). Bratton, William W., Pari Passu and a Distressed Sovereign’s Rational Choices (Draft of Feb. 15, 2004) Buchheit, Lee C. & Gulati, G. Mitu, Exit Consents in Sovereign Bond Exchanges, 48 UCLA L. REV. 59 (2000).

- 92 -

Buchheit, Lee C. & Gulati, G. Mitu, Sovereign Bonds and the Collective Will, 51 EMORY L.J. 1317 (2002) Buchheit, Lee C. & Pam, Jeremiah S., The Pari Passu Clause in Sovereign Debt Instruments, Working Paper, Harvard Law School, Program on International Financial Systems (Draft of December 11, 2003), available at www.law.harvard.edu/programs/pifs/pdfs/buchheitpam.pdf BUCKLEY, ROSS P., EMERGING MARKETS DEBT : AN ANALYSIS OF THE SECONDARY MARKET (1999) Buckley, Ross P., Rescheduling as the Groundwork for Secondary Markets in Sovereign Debt, 26 DENV. J. INT’L L. POL’Y 299 (1998) Bulow, Jeremy & Rogoff, Kenneth, Sovereign Debt: Is to Forgive to Forget?, 79 AM. ECON. REV. 43 (1989) Catan, Thomas & Mulligan, Mark, Argentina in Dollars 155bn Debt Default; Biggest Suspension Of Payments In History; ‘Third Currency’ To Be Created To Kick-Start Economy, FIN. TIMES (London ed.), Dec. 24, 2001, at 1 Catan, Thomas, Hope – and Everything Else – Is Running Out in Buenos Aires: The Economic Crisis Is Swelling the Ranks of the Poor, FIN. TIMES (London ed.), Feb. 2, 2002, at 22 Chan-Lau, Jorge A., Anticipating Credit Events Using Credit Default Swaps, with an Application to Sovereign Debt Crises (IMF Working Paper, May 2003) Choi, Stephen & Gulati, G. Mitu, Why Lawyers Need to Take a Closer Look at Exit Consents, INT’L FIN. L.R., Sept. 2003, at 15 Chun, John H., Note, "Post-Modern" Sovereign Debt Crisis: Did Mexico Need an International Bankruptcy Forum?, 64 FORDHAM L. REV. 2647 (1996) Cohen, Steven M., Comment, Give Me Equity or Give Me Debt: Avoiding a Latin American Debt Revolution, 10 U. PA. J. INT’L BUS. L. 89 (1988) Cole, Daniel H., Debt-Equity Conversions, Debt-for-Nature Swaps, and the Continuing World Debt Crisis, 30 COLUM. J. TRANSNAT’L L. 57 (1992) Conklin, James, The Theory of Sovereign Debt and Spain under Philip II, 106 J. POL. ECON. 483 (1998) Corrales, Javier, Market Reforms, in CONSTRUCTING DEMOCRATIC GOVERNANCE IN LATIN AMERICA 74 (Jorge I. Dominguez & Michael Shifter eds., 2d ed. 2003)

- 93 -

Cymrot, Mark A., Barricades at the IMF: Creating a Municipal Bankruptcy Model for Foreign States, 36 INT’L LAW. 1103 (2002) DOUGLAS G. BAIRD, ROBERT H. GERTNER AND RANDAL C. PICKER, GAME THEORY AND THE LAW (1994) Dufey, Gunter & Rehm, Florian, An Introduction to Credit Derivatives, University of Michigan Business School, Working Paper 00-013 (August 1, 2000) Economics Focus, A Better Way to Go Bust, ECONOMIST, Feb. 1, 2003, at 64 EDWARDS, SEBASTIAN, CRISIS AND REFORM IN LATIN AMERICA: FROM DESPAIR TO HOPE (1995) Eichengreen, Barry & Rühl, Christof, The Bail-In Problem: Systematic Goals, Ad Hoc Means, NBER Working Paper 7653 (April 2000), available at http://www.nber.org/papers/w7653 Eichengreen, Barry, Bailing in the Private Sector: Burden Sharing in International Financial Crisis Management, 23 FLETCHER F. WORLD AFF. 57 (1999) EICHENGREEN, BARRY, FINANCIAL CRISES AND WHAT TO DO ABOUT THEM (2003) Feder, Norman Menachem, Deconstructing Over-the-Counter Derivatives, 2002 COLUM. BUS. L. REV. 677 (2002) Fitch Ratings, Credit Derivatives: A Case of Mixed Signals? (December 4, 2003) Galvis, Sergio J. & Saad, Angel L., Collective Action Clauses: Recent Progress and Challenges Ahead (Draft of Feb. 20, 2004). Galvis, Sergio J., Sovereign Debt Restructurings – The Market Knows Best, 6 INT’L FIN. 145 (2003). Gelpern, Anna, How Collective Action Is Changing Sovereign Debt, INT’L FIN. L.R., May 2003, at 19 Goldman, Samuel E., Mavericks in the Market: The Emerging Problem of Hold-Outs in Sovereign Debt Restructuring, 5 UCLA J. INT’L L. & FOREIGN AFF. 159 (2000) Grossman, Herschel I. & Van Huyck, John B., Sovereign Debt as a Contingent Claim: Excusable Default, Repudiation and Reputation, 78 AM. ECON. REV. 1088 (1988) Gugiatti, Mark & Richards, Anthony, The Use of Collective Action Clauses in New York Law Bonds of Sovereign Borrowers (July 11, 2003), available at: http://ssrn.com/abstract= 443840

- 94 -

Gulati, G. Mitu & Klee, Kenneth N., Sovereign Piracy, 56 BUS. LAW. 635 (2001) HENDERSON, SCHUYLER K., HENDERSON ON DERIVATIVES (2003) Herring, Jane, Credit Derivatives in Emerging Markets: A Product Analysis, in CREDIT DERIVATIVES: APPLICATIONS FOR RISK MANAGEMENT 11 (1998) HUDSON, ALASTAIR, ED., CREDIT DERIVATIVES: LAW, REGULATION AND ACCOUNTING ISSUES (2000) HUDSON, ALASTAIR, THE LAW ON FINANCIAL DERIVATIVES (3rd ed. 2002) Kahler, Miles, Politics and International Debt: Explaining the Crisis, in THE POLITICS OF INTERNATIONAL DEBT 11 (Miles Kahler ed., 1986) Krueger, Anne O., Market Discipline and Public Policy: The Role of the IMF (October 31, 2003) Krueger, Anne, New Approaches to Sovereign Debt Restructuring: An Update on Our Thinking, Address before the Institute of International Economics (April 1, 2002), available at http://www.imf.org/external/np/speeches/2002/040102.htm Lapper, Richard, Creditors Unite to Seek Better Deal on Argentine Debt, FIN. TIMES, Dec. 9, 2003 McKay, Betsy, Moscow Journal: From Worry to Worse – Revisiting a Family in Russia 7 Years Later, WALL ST. J., Sept. 28, 1998, at A1 Miller, Jessica W., Solving the Latin American Sovereign Debt Crisis, 22 U. PA. J. INT’L ECON. L. 677 (2001) Packer, Frank & Suthiphongchai, Chamaree, Sovereign Credit Default Swaps, BIS Q. REV., Dec. 2003, at 79 Partnoy, Frank, ISDA, NASD, CFMA and SDNY: The Four Horsemen of Derivatives Regulation (USD School of Law, Public Law and Legal Theory Working Paper 39) Pollack, Emily R., Assessing the Usage and Effect of Credit Derivatives (April 2003) (Unpublished Paper, Program on International Financial Systems, Harvard Law School) Power, Philip J., Sovereign Debt: The Rise of the Secondary Market and Its Implications For Future Restructurings, 64 FORDHAM L. REV. 2701 (1996) Robert B. Gray, Chairman, International Primary Market Association, Remarks, 2003 AM. SOC’Y INT’L L. PROC. 223

- 95 -

Roe, Mark J., The Voting Prohibition in Bond Workouts, 97 YALE L.J. 232 (1987) Rogoff, Kenneth & Zettelmeyer, Jeromin, Early Ideas on Sovereign Bankruptcy Reorganization: A Survey, IMF Working Paper 02/57 Rule, David, The Credit Derivatives Market: Its Development and Possible Implications for Financial Stability, FIN. STABILITY REV., June 2001, at 117. Sachs, Jeffrey D., Do We Need an International Lender of Last Resort?, Frank D. Graham Lecture, Princeton University 8 (Apr. 20, 1995), available at: www.earthinstitute.columbia. edu/about/director/pubs/intllr.pdf Salmon, Felix, Brazil Goes Off on a CAC Tangent, EUROMONEY, June 2003 Salmon, Felix, Uruguay Closes the Loop, EUROMONEY, May 2003 Scheerer, André, Credit Derivatives: An Overview of Regulatory Initiatives in the U.S. and Europe, 5 FORDHAM J. CORP. & FIN. L. 149 (2000) SCHILLER, ROBERT J., THE NEW FINANCIAL ORDER: RISK IN THE TWENTY-FIRST CENTURY (2003) Schinasi, Garry J., et al., Modern Banking and OTC Derivatives Markets: The Transformation of Global Finance and its Implications for Systemic Risk (IMF Occasional Paper 203) (2000) Schleifer, Andrei, Will the Sovereign Debt Market Survive?, Harvard University, Department of Economics Working Paper (March 2003) Schwarcz, Steven L., Sovereign Debt Restructuring: A Bankruptcy Reorganization Approach, 85 CORNELL L. REV. 956 (2000) Scott, Hal S., A Bankruptcy Procedure for Sovereign Debtors?, 37 INT’L LAW. 103 (2003) SCOTT, HAL S., INTERNATIONAL FINANCE: TRANSACTIONS, POLICY AND REGULATION (11th ed. forthcoming 2004) Silverman, Ronald J. & Deveno, Mark W., Distressed Sovereign Debt, A Creditor’s Perspective, 11 AM. BANKR. INST. L. REV. 179 (2003) Skeel, David A., Jr., Can Majority Voting Provisions Do It All?, 52 EMORY L.J. 417 (2003) STIGLITZ, JOSEPH E., GLOBALIZATION AND ITS DISCONTENTS (2002)

- 96 -

Thomson, Adam, Argentine Bond Default Hits its Pensioners Hardest, FIN. TIMES, March 13-14, 2004, at 2 Tolk, Jeffrey, Moody’s Investor Services, Understanding the Risks in Credit Default Swaps (March 16, 2001) Wheeler, Christopher C. & Attaran, Amir, Declawing the Vulture Funds: Rehabilitation of a Comity Defense in Sovereign Debt Litigation, 39 STAN. J. INT’L L. 253 (2003) White, Michelle J., Sovereigns in Distress: Do They Need Bankruptcy? (2002) BROOKINGS PAPERS ECON. ACTIVITY 287

- 97 -